Skip to main content

Full text of "Orion or the antiquity of the vedas."

See other formats







All rights reserved. 



'* ft '1893. 

. jMA/iana. Vd, xiii. 9, 
. Vd. viiu46, 

[ Seepages 110 7 HL 

Yi'^QiaL. 1 

43 am t AIT- 




SOME explanation may be necessary for the publication 
of an essay on the antiquity of the Vedas by one whose 
professional work lies in a different direction. About four 
years ago, as I was reading the Bhagavad Gita, it occurred 
to me that we might derive important conclusions from fche 
statement of Krishna that ''he was Margashirsha of the 
months. " This led me to inquire into the primitive Vedic 
calendar, and the result of four years' labour is now placed 
before the public. The essay was originally written for 
the Ninth Oriental Congress held in London last year. 
But it was found too large to be inserted in the proceedings 
wherein its summary alone is now included. I have had 
therefore to publish it separately, and in doing so I have 
taken the opportunity of incorporating into it such addi- 
tions, alterations and modifications, as were suggested by 
further thought and discussion. 

The chief res alt of my inquiry would be evident from the 
title of the essay. The high antiquity of the Egyptian civi- 
lization is now generally admitted. But scholars still hesi- 
tate to place the commencement of the Vedic civilization 
earlier than 2400 B.C. I have endeavoured to show 
in the following pages that the traditions recorded in 
the Rigveda unmistakably point to a period not later 
than 4000 B.C., when the vernal equinox was in Orion, or, 
in other words, when the Dog-star ( or the Dog as we have 
it in the Rigveda) commenced the equinoctial year. Many of 
the Vedic texts and legends, quoted in support of this con- 
clusion, have been cited in this connection and also ration- 


ally and intelligently explained for tho first time, tlius 
throwing a considerable light on the legends and rites in 
later Sanskrit works. I have farther tried to show how 
these legends are strikingly corroborated by the legends 
and traditions of Iran and Greece. Perhaps some of this 
corroborative evidence may not be regarded as sufficiently 
conclusive by itself, but in that case I hope it will be borne 
in mind that my conclusions are not based merely upon my- 
thological or philological coincidences, and if some of these 
are disputable, they do not in any way shake the validity 
of the conclusions based on the express texts and references 
scattered over the whole Vedic literature. I wanted to 
collect together all the facts that could possibly throw any 
light upon, or be shown to be connected with the question 
in issue, and if in so doing I have mentioned some "that are 
not as convincing as the others, 1 am sure that they will at 
least be found interesting, and that even after omitting 
them there will be ample evidence to establish the main 
point. I have, therefore, to request my critics not to be 
I prejudiced by such facts, and to examine Mid weigh the 
Iwhole evidence I have adduced in support of my theory 
before they give their judgment upon it, 

I have tried to make the book as little technical as pos- 
sible | but I am afraid that those who are not acquainted 
withtJie Hindu method of computing time may still find ife 
somewhat difficult to follow the argument in a few places. 
If my conclusions come to be accepted and th seeoticl 
edition of the book be called for, these defects ffiay T^e 
removed by adding further explanations in such oftff6& >; - At 
present I have only attempted to give the maiti argument 
on the assumption that the reader is already familiar with 
the method, I may further remark that though I have 


used the astronomical method, yet a comparison with Bent- 
ley's work will show that the present essay is more literary 
than astronomical in its character. In other words, it is the 
Sanskrit scholars who have first of all to decide if my inter- / 
pretations of certain texts are correct, and when this judg- 
ment is once given it is not at all difficult to astronomically 
calculate the exact period of the traditions in the Rigveda. 
I do not mean to say that no knowledge of astronomy is 
necessary to discuss the subject, but on the whole it would 
be readily seen that the question is one more for Sanskrit t 
scholars than for astronomers to decide. ' 

Some scholars may doubt the possibility of deriving so 
important and far-reaching conclusions from the data 
furnished by the hymns of the Rigveda, and some may 
think that I am taking the antiquity of the Vedas too far 
back. Bat fears like these are out of place in a historical 
or scientific inquiry, the sole object of which should be to 
search for and find out the truth, The method of investigation 
followed by me is the same as that adopted by Bentley, 
Colebrooke and other well-known writers on the subject, 
and, in my opinion, the ojaly question that Sanskrit scholars 
have now to decide, is whether I am or am not justified in 
carrying it a step further than my predecessors, indepen- 
dently of any modifications that may be thereby made 
necessary in the existing hypothesis on the subject. 

I have omitted to mention in the essay that a few native 
scholars have tried to ascertain the date of the MaMbh&rata, 
and the Rkn&yana from certain positions of the sun, the 
moon and the planets given in those works. For instance, 
the horoscope of R&ma and the positions of the planets at 
the time of the great civil war, aa found in the Mah&bh&rata, 


are said to point to a period of 5000 or 6000 B. C., and it is 
contended that the Vedas which preceded these works 
mast be older still. Bentley relying on the same data has 
calculated 961 B. C. as the exact date of Rama's birth. 
| This will show how unsafe it is to act upon calculations 
\] based upon such loose statements. Sometimes the accounts 
in the Puranas are themselves conflicting, but even where 
they are or can be made definite any conclusions based on 
them are not only doubtful, but well nigh useless for chro- 
nological purposes, for in the first instance they are open 
to the objection that these works may not have been 
written by eye-witnesses ( the mention of Edshis in the 
Bamayana directly supporting such an assumption ) , and, 
secondly, because it is still more difficult to prove that 
we now possess these books in the form in which they were 
originally written. With regard to the positions of the 
planets at the time of the war given in the Mahabharata, 
the statements are undoubtedly confused ; but apart from 
it, I think that it is almost a gratuitous assumption to hold 
that all of them really give us the positions of the planets 
in the ecliptic and that such positions again refer to the 
fixed and the moveable zodiacal portions of the Nakshatras. 
Perhaps the writers simply intend to mention all auspicious 
or inauspicious positions of the planets in such cases. I 
have therefore avoided all such debatable and doubtful 
points by confining myself solely to the Vedic works, about 
the genuineness of which there can be no" doubt, and using 
the Puranic accounts only to corroborate the results deduced 
from the Vedic texts. According to this view the Maha- 
bharafea war must be placed in the Krittika period, inas- 
much as we are told that Bhishma was waiting for the 
turning of the sun from the winter solstice in the month 



of Magha. The poem* as we now have it, is evidently 
written a long time after this event. 

Lastly, I have to express my obligations to several friends 
for encouraging me to carry on the inquiry and helping me 
in one way or another to complete this essay. My special 
thanks are however due to Dr. Ramkrishna Gopal Bhandar- 
kar, who kindly undertook to explain to me the views of 
German scholars in regard to certain passages from the 
Rigveda, and to Khan Bahadur Dr. Dastur Hoshang Jamasp 
for the ready assistance he gave in supplying information 
contained in the original Parsi sacred books. I am also 
greatly indebted to Prof. Max M tiller for some valuable 
suggestions and critical comments on the etymological 
evidence contained in the essay. I am, however, alone 
responsible for all the views, suggestions, and statements 
made in the following pages. 

With these remarks I leave the book in the hands of 
critics, fully relying upon the saying of the poet 

" The fineness or the darkness of gold is best tested in 
fire. " It is not likely that my other engagements will 
permit me to devote much time to this subject in future ; and 
I shall consider myself well rewarded if the present essay 
does in any way contribute to a fuller and unprejudiced 
discussion of the high antiquity of the Aryan civilization, 
of which our sacred books are the oldest records in the world. 


Poona, October, 1895. 



^ I. Introduction ... * 1 9 


II. Sacrifice alias the Year ... ... 10 31 

III. The Krifctikas 3260 


j l IV. Agrah&yana 61 95 

f V. The Antelope's head 96128 

VI. Orion and. his belt 129156 

I VII. Eibhas and Vrishakapi 157197 

II VIII. Conclusion 198220 


tv Appendix ... ... 221 227 




Researches into the Antipity of the Yedas. 



Importance of ascertaining the A,ge of the Vedas Linguistic method Its 
defects Astronomical method Its difficulties unduly magnified 
Views of European and Native scholars examined . 

THE VEDA is the oldest of ^ the books that we now possess, 
and it is generally admitted " that for a study of man, or if 
you like/ for a study of Aryan humanity, there is nothing in 
the world equal in importance with it. 3 '* There is no other 
book which carries us so near the beginning of the Aryan 
civilization, if not the absolute beginning of all things, as 
maintained by the Hindu theologians ; and the importance 
of ascertaining even approximately the age when the oldest 
of the Vedic Rishis, like the classical Valmiki, may have 
been inspired to unconsciously give utterance to a Vedic 
verse, cannot therefore be overrated- The birth of Gautama 
Buddha, the invasion of Alexander the Great, the inscrip- 
tions of Ashoka, the account of the Chinese travellers, and 
the overthrow of Buddhism and Jaiuism by Bhatta Rumania 
and Shankarach&rya, joined with several other loss important 

* India : what it can teach us ? p. 112. The references through- 
out are to the firat edition of this work. 


s, have served to fix the chronology o the later 
.periods of the Ancient Indian History, But the earlier 
periods of the same still defy all attempts to ascertain their 
chronology ; and the earliest of them all, so important to the 
*' trne student of mankind," the period of the Rigveda^ is 
still the snbject of vague and uncertain speculations. Qan 
we or can we not ascertain the age of the Vedas ? This is 
a question which has baffled the ingenuity of many an. 
ancient and a modern scholar, and though I have ventured to 
write on the subject I cannot claim to have finally solved 
this important problem in all its bearings. I only wish, to 
.place before the public tte result of my researches in this 
direction and leave it to scholars to decide if it throws any 
^additional light on the earliest periods of the Aryan civili- 

But before I proceed fco state nay views, it may be useful 
4o briefly examine the methods by which Oriental scholars, 
have hitherto attempted to solve the question as to the age 
and character of the Vetias. Brof . Max Miiller divides the 
ic literature into four periods the Chhandas, Mantra., 
and Sutra; and as each period presupposes the 
jpraeediag, while the last or the Siitra period is prior, < if 
to the origin, at least k> the spreading and political 
.aaoeitdiuftey of Bnddhism " in the fonrtk -century before 
Christy thai* Le&mod. scholar, by assigning two hundred 
jears for each period arrives at abont 1200 B, 0,, as the, 
dftfo^ at which we may suppose the Vedic hymns to 
JIHT feeea e^mposedA This, for convenience, may Tbe 

* See Man MtEef's Irt Ed. of Big. VoL IV., Pref. pp. ^ vii 
TJiie prc& k,Aho.pri&ted-aB a separate pamphlet under th* ^itte 
** ATCiemt Hindu AstecmOT^ aad Chronology." In the SWOBC! 

*t^uik'^f AeMigve^ ^ie prcfocos-ifi the irst edition arfe reprinted 

wgettieF mi tke fegiaiiiag erf .the J0iirth 


called the literary or the linguistic .method of ascertaining 
the age of the Vedas. A little consideration will, however, 
at once disclose the weak points in such arbitrary calcula- 
tions. There are different opinions as to the division of r 
the Vedic literature ; some scholars holding that the Chhan- 
das and Mantra is one period, though a long- one. But 
granting that the Vedic literature admits of a four-fold' 
division, the question of the duration of. each period", is still* 
involved in uncertainty and, considering the fact that each* 
period might run into and overlap the ofcher to- a certain^ 
extent, it becomes extremely difficult to assign even the- 
minimum chronological limits to the -different periods. The- 
method may, indeed; be used with advantage to show that 
the Vedas could not have been composed later than a cer- 
tain period; but it helps little in eve a- approximately fixing 
the correct age of the Vedaa. Prof. M&x Mutter himself 
admits* that the limit of 200 years carjube assigned to ea,ch 
period only under the supposition that during ther- early 
periods of history the growth of the^human mind' was more- 
luxuriant than in later times-; while the late Dr}. Haug,, 
following the same method, fixed the very commencement 
of the Vedic literature between 2400-2000 B. C.,f by' 
assigning- about 500 years to each period^, on the analogy of 
similar periods in tire Chinese literature. It is- therefore- 
evident that this method of calculation^ howsoever valuable- 

* Pref. to Rig, Vol. IV., p. vii. 

t Introduction to the Aitareya Brahmana, p. 48. Prof. Whitney 
thinks that the hymns. may have been sung as early as 2000 Bi G. 
Fide Intro* to his Sanskrit Gramtnaiyp, xiii. For a summary of 
opinions of different scholars on this point see Kaegi's- lligveda 
by ArroWsmithj p^ 110,: note 39. The highest antiqjxity 


it may be in checking the results arrived at "by other me- 
thods, is, when taken by itself, most vague and uncertain* 
A farther study of the different periods of the Vedic 
literature and its comparison with other ancient literatures 
nrigHt hereafter help us to ascertain the duration of each 
period a little more accurately.* But I think we cannot 
expect, by this method alone, to be ever in a position to fix 
with any approach to certainty the correct age of the Vedas. 
Prof. Max Muller considers 200 years to be the minimum 
duration of each period* while Dr. Haug and Prof. Wilson 
thooght that a period of 500 years was not too long for the 
purpose ;f and I believe there is hardly any inherent impro- 
bability if a third scholar proposes to extend the duration 
of eaeh of these periods up to something like 1000 years. 
In t3*e face of this uncertainty we must try to find out other 
means for ascertaining the correct age of the Vedas* 

The Vedas, the Brahmanas and the Sutras contain numer- 
ous allusions and references to astronomical facts, and it was 
believed that we might be able to ascertain from them the 
a*ge of the oldest literary relic of the Aryan race. But 
somehow or other the attempts of scholars to fix the age of 
Ihe Vedas by what may be called the astronomical method^ 
have not yet met with the expected success. Unfortunately 
for us, alt the Sanskrit astronomical works that we now 

* IB a p&per submitted to the Niath Oriental Congress, Mr-. 
Dhmva has recently examined the whole Vedic literature with a 
view to ascertain its chronology* and he arrives at the conclusion 
that the ctnnttion assigned to the several periods of the Vedic 
literature by Prof. Max Mailer is toe short, and that "without 
making any guesses at numbers of years or centuries** we should 
at pfesent he content with arranging the Vedic literature 
after the manner of the Geological strata or periods. 

t See Ait. Br. Intr., p. 48 ; also Pref. to Big. VoL IV^ p 


possess, except perhaps the Vedanga Jyotishaj belong to the 
later period of Sanskrit literature, when the Greek influence 
is perceptible in. all its mathematical works. The different 
methods of astronomical calculations given in these works-, 
the various eras that were established in India after Shali- 
v&hana or Vikrama, the introduction of the Barhaspatya 
cycle, and the adoption of the Greek division of the Zodiac-, 
make it extremely difficult to correctly interpret the astrono- 
mical references in the later works ; while the confusion, 
caused by the supposed absence of any definite statement as 
to the character of the year and the cycle mentioned in the 
Vedic works, renders it a hard task to deduce a consistent 
theory out of the various but stray references to astrono- 
mical facts in the Vedic literature. Take for instance the 
question of the commencement of the year in the Vedic 
calendar. There are grounds to hold that the ancient Aryas 
commenced their year either with spring or with autumn, at 
the equinoxes or at the solstices ;* while the later astrono- 
mical works and systems furnish us with facts which go to 
prove that the year, in the different parts of India, com* 
monced with almost all the different months of the year 
K&rtika, M&rgashJrslia,t AsMdha, Chaitra and Bhadrapada. 
The discussion as to the number of the Nakshatras and 
different opinions as to their origin have further complicated 
the problem ; while doubts have been raised as to the 
capacity of the Brfthmaus in 3200 B. C. to make observa- 
tions of solstitial points with astronomical accuracy,:]: I shall 

* See infra Chap. II. 

t Whitney's Surya SiddMnta xiv., 16, n. 

J Pref. to Rig. Vol. IV,, p. xxix. It is very difficult to under- 
stand on what grounds this assertion is made. Ancient Vedic bards 
had no mathematical instruments, but still they could have easily 
marked when day and night became equal in length. 



have to examine hereafter how far some of these objections 
are tenable. For the present it is sufficient to state that in 
consequence o such doubts and objections, definite obser- 
vations or allusions to astronomical events in the earliest 
works have been looked upon with suspicion by t a good 
many Oriental scholars, while some have "even condemned 
the astronomical method as inaccurate and conjectural.* 
It is, however, admitted that "if the astronomical data on 
which conclusions as to the age of the Veda have been 
built implied all that they were represented to imply, the 
earliest periods of Yedic- poetry will have to be rearranged.' 7 t 
It appears to me that scholars have erred too- much on the 
mde of overcaujtiousness in condemning this method. I do 
*mot mean to say that there are no difficulties ; but sufficient 
eare does not appear to have been taken to always keep in 
'Tie w the main point of the- inquiry, by separating it from 
the mass of irrelevant matter, with which, in some cases, 
it becomes unavoidably mixed up,.- Some 0f Bentley's 
speculations, for instance, are indeed ingenious and sug- 
gestive, but he relies too much upon Puranic traditions, 
mere etymological speculations and his own calculations 
b^sed thereon, instead of trying to find out whether there is 
anything in the earlier works to corroborate or support these 
traditions; On the other hand, Prof. WebWs Essay, 
wkidlt, as a collection of astronomical allusions and referen- 
ces in the Vedic -literature, is extremely valuable, is takeh 
lip by the controversy as to the origin of the- Kakshatrate 
raised by M, Biot ; and the same thing may be said of 
"Prof. Whitney's contributions on the subject.J . Various- 

* $ee Weber's History of Indian Literature, p. 2, aote. 
f , Pref. to JB% Yol. IV., p. Uxi. 

J See his essay on the Hindu and Chinese systenas* f 


other questions, such as whether the Vedic cycle comprised 
five or six years, how and when the intercalary days or 
months were inserted to make the lunar correspond with 
the solar year, have also caused the attention of scholars to 
be diverted from the broad astronomical facts and observa- 
tions to be found recorded in the Vedic literature ; and as a 
consequence we find that while the questions as to the 
original number of the Nakshatras and as to whether the 
Chinese borrowed them from the Hindus or vies versa, are 
so ably discussed, no systematic attempt has yet been made 
to trace back the astronomical references in the - later 
works to the Sauhitas, and to fully examine their bearing 
on the question of the age and character of the Vedas. On 
the contrary, Prof. Weber asks us to reconcile ourselves to 
the fact that any such search will, as a general rule, bo 
absolutely fruitless !"* In the following pages I have endea- 
voured to shew that we need not be so much disappointed. 
In my opinion there is ample evidence direct and circum- 
stantial in the earliest of the Sanhitas, to fully establish 
the high antiquity assigned to the Indian literature on 
geographical and historical grounds, f I base my opinion 
mainly upon references to be found in the early Vedic 
works, the Sanhitas and the Br&hrnanas, and especially in 
the earliest of these, the Rigveda. For though later works 
may sppaetimes give the same traditions and references/ yefe 
any in&;rep.ce which is based upon them is likely to be re- 
garded with more or less suspicion, unless we can show 

* Weber's History of Indian Literature, p. 7. 
t It is on these grounds that Prof. Weber believes that the be- 
ginnings of the Indian Literature " may perhaps he traced back 
to the time when tbe Indo- Aryans still dwelt together with 
l*4ttBa*'Aryan&. ".Hist. Ind. JLit,, p, 5. 


something in the 'earliest works themselves to justify that 
inference. Where the Sanhitas and the Brahmanas directly 
speak of the actual state of things in their time, there is, of 
course, no ground to disbelieve the same, but I think that 
even the traditions recorded in these works are more reliable 
than those in later works, for the simple reason that those 
traditions are there found in their purest form. Later 
works may indeed be used to supply confirmatory evidence, 
where such is available; but our conclusions must in the 
^ main be based on the internal evidence supplied by the 
ij Vedic works alone. Several Indian astronomers have 
' worked more or less on the lines here indicated, but their 
labours in this direction have not unfortunately received 
the attention they deserve. The late Krishna SMstri 
Godbole published his views on the antiquity of the Vedas 
in the second and third Volumes of the J'heosop'hist,* and 
though he has failed to correctly interpret some astrono- 
mical allusions in the Vedic works, yet there is much that 
is suggestive and valuable in his essay. The 'late Prof. 
X. L. Ohhatre also appears to have held similar views on 
&e -subject, but he has not published them, so far as I 
fcb*w, in a systematic form. My friend Mr. Shankara 
Bilkriahrta Dikshit, who has written a prize essay in. 
Marlthi on the history of Hindu Astronomy, and who has 
succeeded in correctly interpreting more verses in the 
Vedanga Jyotishaf than any other scholar has hitherto 

* Also published, as a separate pamphlet. 

t Mr. Dikshit would do well to publish an English translation 
of at least the Chapter on .Vcdanga Jyofcisha in his essay. He has 
undoubtedly made a great advance over Weber and Thebaut in the 
correct interpretation of the treatise. 



done, has also discussed this question in his essay,, which I 
was allowed to read in MS. through his kindness. I am 
indebted to these scho]ars for some of the facts and argu- 
ments set forth in the following pages, and the present essay 
may, I think, be regarded as greatly developing, if not 
completing, the theory started by them. 



Primitive calendar co-eval with the sacrificial system Prajapatit=Yaj Da 
= Scmwabtara Civil or s&uana days Sva>wi and lunar months 
Lunar and solar years Intercalary days and month, in Vedic times- 
Solar year was siderial and not tropical Old beginning of the year 
and the sacrifice. The Yishuv&n day Vernal equinox and winter 
solstice Uttar&yaBa and Datohin&yaaaa Bevay&na and Pitriyna 
Their original meaning Bhiskar4cMrya r s mistake about the day of 
theBems The two year beginningB were subsequently utilised for 
different purposes. 

IT is necessary, in the first place, to see what contri- 
vances were adopted by the ancient Aryas for the measure- 
ment and division of time. The present Indian system has 
been thus described by Professor Whitney in his notes to 
the Siirya Siddhanta (r. 13, notes) : 

fc In the ordinary reckoning of time* these elements are 
u variously combined. Throughout Southern India (see 
*' Warren's Kala Sankalita, Madras, 1825, p. 4, etc.), the 
M year and month made use of are the solar, and the day the 
** 3ivi} ; Hie beginning of each month and year being 
* c esotmted, in practice, from the sunrise nearest to the 
* moment of their actual commencement. In all Northern 
* c India the year is luni-solar ; the month is lunar and is 
C divided into both lunar and civil days ; the year is com- 
** posed of a viable number of months, either twelve or 
" &irteett> "beginning always with the lunar month, of which 
QJB commEicemanf next precedes the true commencement 
*"* 0f& Cereal year* But underneath this division* 

ear into twelve so&tf 


" is likewise kept up, and to maintain tlie concurrence of 
" the civil and lunar days, and the lunar and solar months, 
" is a process of great complexity, into the details of which 
<f we need not enter here.'* 

But the complications here referred to are evidently the 
growth of later times. The four ways of reckoning time, 
the Sftvana, the Chandra, the Nakshatra and the Saura, are 
not all referred to in the early works, and even in later days 
all these measures of time do not appear to have been fully 
and systematically utilised. There is, as I have said before, 
no early work extant on Vedic calendar, except the small 
tract on Jyotisha, and our information about the oldest 
calendar must, therefore, be gathered either from stray 
references in the Vedic works or from the early traditions 
or practices recorded in the old sacrificial literature of 
India. There are several sacrificial hymns in , the Rigveda, 
which show that the sacrificial ceremonies must then have 
been considerably developed ; and as no sacrificial system 
could be developed without the knowledge of months, sea- 
sons, and the year, it will not be too much to presume that 
in Vedic times there must have existed a calendar to regu- 
late the sacrifices. It is difficult to determine the exact 
nature of this calendar, but a study of the sacrificial litera- 
ture would show that the phases of the moon, the changes 
in the seasons, and the southern and northern courses of 
the sun were the principal land-marks in the measurement 
of time in those early days. What is still more interesting, 
however, is that the leading features in the early sacrifices 
are the same as those in the year. The late Dr. Haug, in 
his introduction to the Aitareya Br&hmana,, has observed 
that "the satras, which lasted for one year, were nothing but 
81$. imitation of the sun's *yearly course. They were divided 


into two distinct parts, each consisting of six months of 80 
days each. In the midst of both was the Vishuvdn, i. e., 
the equator or the central day, catting the whole satra into 
two halves."* This clearly shows that the ancient Rishis 
prepared their calendar mainly for sacrificial purposes, and 
the performance of various sacrifices facilitated, in its turn, 
the keeping up of the calendar. Offerings were made every 
morning and evening, on every new and full moon, and at 
the commencement of every season and ay ana. f When 
this course of sacrifices was thus completed, it was naturally 
found that the year also had run its course, and the sacri- 
!e&ai the year, therefore, seem to have early become 
synonymous terms. There are many passages in the 
Br^bmanas and Sanhit&s, where Samvatsara and Yajna are 
declared to be convertible terms, J and no other theory has 
yet been suggested on which this may be accounted for. I 
am therefore inclined to believe that the Vedic Rishis kept 
up their calendar by performing the corresponding round 
of sacrifices on the sacred fire that constantly burnt in their 
houses, like the fire of the Parsi priest in modern times. 
e numerous sacrificial details, which we find so fully 
in the Brahmanas, might be later innovations, 

* AH. Br. Intiv P* 48. 

f Cl Ba&clMyaiia Sutras, ii. 4.23, which describes the continuous 
rani of sacriieies as follows : 

I Also com- 

Mann iv. 25-26, ai^ Yajnavalkya i. 125. 
J See Ait. Br. S. 17* wMch says ^j^c^i: !y^nm^ ; \ 

; Biat&patha Br. xL 1. 1. 1 ; 2. f. 1. IB Taitt. 
n. i. 7. % -tiL-5. 7. 4 we have ^frl H14H^ amd again 



but the main idea of the yearly sacrifice appears to be an 
old one. The etymology of the word ritvij ( ritu+yaj^= 
season sacrificer) shows that even in the oldest days there 
existed a certain correspondence between the sacrifices and 
the seasons, and what is true of the seasons is true of the 
year which according to one derivation of samvatsara (vas 
= to dwell ) is nothing but a period where seasons dwell, or 
a cycle of seasons.* The priests were not only the sacri- 
ficers of the community, but were also its time-keepers, t 
and these two functions they appear to have blended into 
one by assigning the commencement of the several sacrifices 
to the leading days of the year, on the natural ground that if 
the sacrifices were- to be performed they must be performed 
on the principal days of the year. J Some scholars have 
suggested that the yearly satras might have been subse- 
quently invented by the priests. But the hypothesis derives 
little support from the oldest records and traditions of all 
the sections of the Aryan race. Without a yearly satra 

* Of. Bhanu Dikshita's Com. on Amara i. 4. 20. Dr. Schrader, 
in his Prehistoric Antiquities of the Aryan Peoples, Part iv., Ch. vi. 
(p. 305), also makes a similar observation. He holds, on philological 
grounds, that the conception of the year was already formed in the 
primeval period by combining into one whole the conception of 
winter and summer, which he believes to be the two primeval 

f " In Rome the care of the calendar was considered a religious 
function, and it had from earliest times been* placed in the hands 
of the pontiffs/' Lewis's Historical Survey of the Astronomy of the 
Ancients, p. 24. 

J " Plato states that the months and years are regulated in 
order that the sacrifices and festivals may correspond with the 
natural seasons ; and Cicero remarks that the system of intercala- 
tioa was introduced with this object." Lewis's His; Aste Aac^ p. 19. 



regularly kept up, a Vedic Rishi could hardly have been 
able to ascertain and measure -the course of time in the way 
he did. When better contrivances were subsequently 
discovered the sacrifices might naturally become divested of 
their time-keeping function and the differentiation so caused 
might have ultimately led to an independent development 
of both fche sacrifices and the calendar. It is to this stage 
that we must assign the introduction of the numerous details 
of the yearly sacrifice mentioned in later works ; and thus 
understood, the idea of a sacrifice extending over the whole 
year, may be safely supposed to have originated in the oldest 
days of the history of the Aryan race."* In fact, it may be 
regarded as coeval with, if not antecedent to, the very 
beginning of the calendar itself. 

We have now to examine the principal parts of the year, 
alias the sacrifice. The sdvana or the civil day appears to 
have been, as its etymology shows, t selected in such cases 
as the natural unit of time. 30 such days made a month 
and 12 such months or 360 sdvana days made a year. $ 
Comparative Philology, however, shews that the names 

* Comparative Philology also points to the same conclusion ; 
, Cf. Sanskrit yaj 9 Zend ya, Greek^os. It is well-known that the 
sacrificial system obtained amongst the Greeks, the Romans and 
the Iranians. 

t Sdvana is derived from su to sacrifice, and means literally a 
sacrificial day. 

I Ait. Br. ii. 17; Taitt. San. ii. 5. 8. 3; Rig, i. 164. 48. Prof. 
Whitney ( Sur. Sid. i. 13,n ) observes, "The civil (s&vana) <% is 
the natural day .... A month of 30 and a year of &60 days are 
supposed to have formed the basis of the earliest Hin4a Chronology, 
an intercalary month being, added once in five years./' 

IT.] SACRIFICE alias THE YEAR. ' 15 

for the month and the moon coincide, with occasional small 
differences of suffix,* inmost of the Indo-European languages, 
and we may therefore conclude that in the primitive Aryan 
times the month was determined by the moon. Now a 
month of thirty civil or s&vana days cannot correspond 
with a lunar synodical month, and the Brahmavadins liad 
therefore to omit a day in some of the sdvana months to 
secure the concurrence of the civil and the lunar months. t 
The year of 360 sdvana days was thus practically reduced 
to a lunar year of 354 civil days or 360 tithis. But a further 
correction was necessary to adjust the lunar with the solar 
reckoning of time. The zodiac was not yet divided into 
twelve equal parts , and the solar month, as we now under- 
stand it, was unknown. The commencement of the cycle of 
seasons was, therefore, the only means to correct the calendar, 
and the ancient Aryas appeared to have early hit upon the 
device of the intercalary days or month for that purpose* 
There are many passages in the Taittirlya and V&jasaneyi 
Sanhit&s and also one in the Rigveda$ wherein the inter- 
calary month is mentioned, and though opinions may differ 
as to when and how it was inserted, we may, for the purpose 
of our present inquiry, regard it as undisputed that in the 
old 7edic days means were devised and adopted to secure 

* See Dr. Schrader's Prehistoric Antiquities of the Aryan 
Peoples, Part iv., Chap. vi. Translation by Jevons, p. 300. Also 
Max Mullet's Biographies of Words, p. 193.- 

t gR^r^TT^ % ^S^fTsffcrflf 4r^Nf^ ^rW^lf^T: I Taitt. San. vii. 
5. 7. 1 4 , and Tanclya Br. v. 10. See also Kala-Madhava Chap, on 
Month, Cal Ed", p. 63. 

J Taitfc: SatL i. 4. 14 ; Vaj. San. 7. 30 ; Rig. i. 25, 8. As regards 
the twelve hallowed (intercalary) nights Cf. Rig. iv. 33. 7; Atha. 
Veda iv. 11. 11; Taitt. Br. i. 1. 9. 10. 


the correspondence of the lunar with the solar year. The 
occurrence of the twelve hallowed nights amongst the 
Teutons points to the same conclusion. They were in fact 
the supplementary days (366-354=12) required to balance 
the lunar with the solar year, a period when the Ribhus, or 
the genii of the seasons, slackened their course and enjoyed 
the hospitality of the sun after toiling for a whole year (Rig. 
i. 33. 7.),* and when Prajapati, the God of sacrifices, after 
finishing the old year's sacrifice, prepared himself for the 
new year's work (Atharva Veda iv. 11. 1L). The sacrificial 
literature of India still preserves the memory of these days 
by ordaining that a person wishing to perform a yearly 
sacrifice should devote ] 2 days ( dvddasMJia ) before its 
commencement to the preparatory rites. These facts, in 
my opinion, conclusively establish that the primitive Aryans 
had solved the problem involved in balancing the solar with 
the lunar year. There may be some doubt as to whether 
the concurrence of the two years was at first secured by 
intercalating twelve days at the end of every lunar year, or 
whether the days were allowed to accumulate until an 
intercalary month could be inserted. The former appears to 
have been the older method, especially as it has been utilised 
and retained in the performance of yearly sacrifices ; but 
whichsoever may be the older method., one thing is certain, 
that primitive Aryas had contrived means for adjusting the 
lunar with the solar year. Prof. Weber and Dr. Schraderf 
appear to doubt the conclusion on the sole ground that we 

* See Zimmer's Mfe in Ancient India, p. 366 ; Kaegi's Blgveda 
(translation by Arrowsmith), pp. 20, 37. 

f Seelndische Studien* xviii. ^24, and Dr. Schrade/s observa- 
tions thereon in his Prehistoric Antiquities of Aryan Peoples, Part 
iv,, Chap, vi., pp. 308-10. 




cannot suppose the primitive Aryans to have so far advanced 
in civilization as to correctly comprehend such, problems. 
This means that we must refuse to draw legitimate inferences 
from plain facts when such inferences conflict with our 
preconceived notions about the primitive Aryan civilization. 
I am not disposed to follow this method, nor do I think that 
people, who knew and worked in metals, made clothing of 
wool, constructed boats, built nouses and chariots, performed 
sacrifices, and had made some advance in agriculture,* were 
incapable of ascertaining - the solar and the lunar year. 
They could not have determined it correct to a fraction of a 
second as modern astronomers have done; but a rough 
practical estimate was, certainly, not beyond their powers 
of comprehension. Dr. Schrader has himself observed that 
the conception of the year in the primeval period was 
formed by combining the conceptions of the seasons.f If 
so, it would not be difficult, even for these primitive Aryans, 
to perceive that the period of twelve full moons fell short 
of their seasonal year by twelve days. Dr. Schrader again 
forgets the fact that it is more convenient, and hence easier 
and more natural, to make the year begin with a particular 
season or a fixed position of the sun in the heavens, than 
to have an ever-varying measure of time like the lunar year. 
Lewis, in his Historical Survey of the Astronomy of the 
Ancients, quotea Geminus to shew that ce the system 
pursued by the ancient Greeks was to determine their 
months by the moon and their years by the sun," J and 

* For a short summary of the primitive Aryan civilization, see 
Peile's Primer of Philology, pp. 66, 67 5 also Kaegi's Rigveda, 
translated by Arrowsmith, pp. 11-20. 

f See Preh, Ant..Ary. Peoples translated by Jevons, p. 305. 

J Le\vi y Hist, Surv. Astron* Anc., p. 18, 


this appears to me to have been tie system in force in the 
Indo-Grerrnanie, or at any rate in the primitive Vedic period. 
There is no other conclusion that we can fairly draw from 
the facts and passages noted above. 

There is, however, a farther question, as to whether the 
solar year, with reference to which these corrections were 
made^ was tropical or sidereal. It is true that the great 
object of the calendar was to ascertain the proper time of 
the seasons. But the change in the seasons consequent 
npon the precession of the equinoxes is so exceedingly 
minute as to become appreciable only after hundreds of 
years, and it is more probable than not that it must have 
escaped the notice of the early observers of the heavens, 
whose only method of determining the position of the 
sun in the ecliptic was to observe every morning the fixed 
stars nearest that luminary.* Under such a system the 
year would naturally be said to be complete when the sun 
returned to the same fixed star. Prof. Whitney has pointed 
out that the same system is followed in the Surya Siddbanta, 
thoQgh the motion of the equinoxes was then discovered.t 
It is, therefore, natural to presume that the early Vedic 
priests were ignorant of the motion of the equinoxes. Eo 
early work makes any mention of or refers to it either 
expressly or otherwise ; and the solar year mentioned in 

Taitt* Br. i. 5. 2. 1 ; ^u^* s^pT tiS^egq'fcih'eHii 1 sppr 

I ,3ftf ^fcf i This is still tfMtteti at 


t Snr. Sid. i. 13. n. " It is* however, not the tropial solar year 
whieb we employ, feut tfoe sidereal, no account being miaJe 0f the 
precession of the ' 


the Vedic works must, therefore, be considered as sidereal 
and not tropical. This would necessitate a change in the 
beginning of the year, every two thousand years or so, 
to make it correspond with the cycle of natural seasons, 
and the fact that such changes were introduced twice or 
thrice is a further proof of the old year" being a sidereal 
one.* The difference between the sidereal and the tropical 
year is 20*4 minutes, which causes the seasons to fall back 
nearly one lunar month in about every two thousand years, 
if the sidereal solar year be taken as the standard of measure- 
ment. When these changes and corrections came to be 
noticed for the first time, they must have created a great 
surprise, and it was not till after one or two adjustments 
on this account were made that their true reason, the motion 
of the equinoxes, could have been discovered, Garga tells 
us that if the sun were to turn to the north without reaching 
Dhanihtha, f it foretold great calamity, and I am disposed 
to put a similar interpretation upon the story o Praj&pati 
alias Yajna alias the year, who, contra-ry to all expectations, 
moved backwards to his daughter Rohini. $ But as I wish 
to examine the tradition more fully hereafter, it is not neces- 
sary to dilate on the point here. My object at present 
is to show that the Vedic solar year was sidereal and 
not tropical, and what has been said above is, I believe, 

* The Krittikas once headed the list of the Nakshatras, which 
now begins with Ashvini. Other changes are discussed in the 
following chapters of this work. 

t Garga quoted by Bhattotpala on Brihat. San. iii. 1 : 

J Ait. Br. iii.. 33. The passage is discussed in this light further 
on ia Chapter VIIL See also Shat, Br. i. 7. 4. 1. 


sufficient to justify sncli a presumption, at least for tlae 
present, though it may afterwards be either retained or 
discarded, according as it tallies or jars with other facts. 

Opinions differ as to whether the lunar month began wifcli 
the full or the new moon,* and whether the original number 
of Natshatras was 27 or 28. f But I pass over these and 
similar other points as not very relevant to my purpose, 
and take up next the question of the commencement of 
the year. I have already stated that the sacrifice and 
the year were treated as synonymous in old days, and we 
may, therefore, naturally expect to find that the beginning 1 
of the one was also the beginning of the other. The Ved&nga 
f Jyotisha makes the year commence with the winter solstice, 
and there are passages, in the Shrauta Sutras which lay 
down that the annual sacrifices like t/avd-m-ayana^ should 
be begun at the same time.$ A tradition has also been 
recorded by Jaimini and others that all Deva ceremonies 
should be performed^ only during the Uttarayana; and the 
"Ottarayana, according to the several Jyotisha works,]] is the 
period of the year from the winter to the summer solstice, 

e * SeeK&Ia Madhava, Chapter on Month, Gal. Ed^ p. 63; <frf *TPE4- 
^ *gtT: 3vHTOT ij&FZ I We can thus explain why the full moon 
night of a, month was described as theirs* night of the year. See infra. 

t ^*e &> fcg* Vol. IV,, and Whitney's Essay on the Hindu 
aiwi Chinese Astecisms. 

$ S*e Ved. Jy./ 5 ; Ashvalayana Shr. Su. i 2. 14. 1 ; ii. 2. 14, 
S ftBd 22; MuSto, So- Y. 1. L " 

mmto^ Barsluma, TI. 8^ ^. Ashvalayaua, Gr. Sa. i 4. 1, 

' " in Kala Madhava, 

Chapter oo j%W% vattHA^'pl 57 t bat from 'the 


that is, from the time when the sun turns towards the north 
till it returns towards the south. This leads one to suppose 
that the winter solstice was the beginning of the year and 
also of the Uttarayana at the time when the annual sacrifices 
were established, and therefore in the old Vedic days. 
But a closer consideration of the ceremonies performed in 
the yearly satras will show that the winter solstice could 
not have been the original beginning of these satras. The 
middle day of the annual satra is called the Vishuvan day, 
and it is expressly stated that this central day divides 
the satra into two equal halves, in the same way as the 
. Vishuvan or the equinoctial day divides the year.* The 
satra was thus the imitation of the year in every respect, 
and originally it. must have corresponded exactly wifch the 
course of the year. Now, as Vishuvan literally means the 
time when day and night are of equal length, if we suppose 
the year to have at the time commenced with the winter 
solstice, the Vishuvan or the equinoctial day could never 
have been its central day; and the middle day of the satra 
would correspond, not with the equinoctial, as it should, 
but with the summer solstice. It might be urged that 
Vishuvan as referring to the satra should be supposed to be 
used in a secondary sense. But this does not solve the 
difficulty. It presupposes that Vishuvdm, must have been 
used at one time in the primary sense (i. e. ? denoting the 
time when day and night are equal); and if in its primary 
sense it was not used with reference to the satm, it must 
| have been so used at least with reference to the year. But 

if Vishuvan was thus the centra] day of the year, the year 
must have once commenced with the equinoxes. The word 
uttardyanaiB again susceptible of two interpretations. It 

* Ait. Br. i?. 22 ? Taitt. Br. i. 2. 3. 1 - T&n. Br, iv. 7. 1. 




the summer solstice.* But notwithstanding their high 
authority it will be found that their interpretation, though 
in consonance with the later astronomical views, is directly 
opposed to the passages in the Vedic works. In the 
Taittinya Sanhita vi. 5. 3, we are told "the sun/ theref ore, 
goes by the south for six months and six by the north. " But 
this does not help us in ascertaining the correct meaning 
of the phrase " by the north." As it stands it may mean 
either the solstitial or the equinoctial six months. We must 
therefore look for another passage, and this we find in the 
Bhatapatha Brahmana (ii. ]. 3,. 1-3), wherein describing 
the two aforesaid paths it lays down, in distinct terms that 
Vasanta, Grishnm and Varsha are the seasons of the Devas ; 
Sharad, Hemanta and Shishira those of the Pitris the in- 
creasing fortnight is of the Devas ; the decreasing one of the 
Pitris: the day is .of the Devas; the night of the Pitris: 
again the fir st part of the day is of the Devas ; the latter of 
the Pitris ...... When he (the sun) turns to the north, he is 

amongst the Devas and protects them ; when he turns to 
the south he is amongst the Pitris and protects them."t 
This removes all doubts as to what we are to understand by 
devaydna, devapaiha, or devaloTca and uttardyona as con- 
nected with it. The Brihadaranyaka Upanishad is a part of 
the Shatapatha Brahmana, and we shall not be violating any 
rule of interpretation if we interpret the passage in the one 

* Shankaracharya is not explicit yet his referenc to the death 
of Bhishma shows that .he takes the same view. Anandagifi on 
Prashnopanishad i. 9, says 

As the passage is important I give it here in full : 





in tlie light of a similar passage ia the otliei\ Now if Va~ 
santa (spring), Grlshnia (summer) and Varsha (rains) were 
tlie seasons of the Dev&s and the sun moved amongst the 
Devas when he turned to the north, it is impossible to 'main- 
tain that the Devayana or the Uttarayaua ever commenced 
with the winter solstice, for in neither hemisphere the winter 
solstice marks tli beginning of spring,, tlie first of the 
Deva seasons* The seasons in Central A&ia and India differ. 
Thus the rains in India commence about or after the 
summer solstice, while in the plains of Asia the season 
occurs about the autumnal equinox:. But in neither cas-e 
Vasanta (spring) commences with, the winter solstice or 
Varsha (rains) ends at the summer solstice. Wo rawst; 
therefore kold that Decay&na in those days was understood 
to extend over the six months of the year, which comprised; 
the three seasons of spring, summer, and rains, i. e., froml 
the vernal to the autumnal equinox, whoa the sun, was in I 
the northern hemisphere or -to the north of the equator J 
This shows farther that the oldest order of seasons did noil 
place Varrfha (rains) at the summer solstice, when the chief 
Indian monsoon commences; but at the autumnal equinox. 
The winter solstice, according to this order, falls in the 
middle of Hemanta. In the modern astronomical works, 
the winter solstice is, however, placed at the end and not in the 
middle of Hemanta, while the vernal equinox is said to fall 
in the middle of Vasanta. When the Vedic Aryas became 
settled in India, such a change in the old order of seasons 
was necessary to make them correspond with the real 
aspect of nature. But it is difficult to determine exactly 
when this change was made.* The old order o seasons 

* Sec Ziinrner's Life in Ancient India, p. 371 . Kaogi's Rigveda, 
f>. 116, note 68-. 


given in the passage above quoted, however, clearly states 

jfchat Vasanta in old days commenced with the vernal 

jequinox. We can now understand why Vasanta has been 

I spoken of as the first season and why the Nakshatras have 

been divided into two groups called the Deva Nakshatras 

and the Yama Nakshatras.* I am aware of the theory 

which attempts to explain away the passages above cited 

as metaphorical to avoid the appearance of superstition. f 

But the method is neither sound nor necessary. The path 

of the Devas and the path of the Pitris are several times 

referred to in the Rigveda, and though we migkt sppose the j 

Brahmavldins to have developed the two ideas to thei^r '^Sf 

extent, it cannot be denied that the original ideais 

an old one, suggested by the passage of the sun in the 

northern and southern hemispheres. 

In the absence of anything to the contrary we niighfe 
j therefore take it as established that in the early Vedic days 
| the year began when the sun was in the vernal equinox ; 
and as the sun then passed from the south to the north ot 
the equator it was also the commencement of his northern 
In other words, the Uttariyaaa (if soch a word 
was then used), Vasanta, the year and the tfatras all com- 
menced together at the vernal equinox. The autumnal 
equinox, witiefa came after the rains was the central day of 
the year ; and tie latter half of the year was named the Pitri- 
yina or what we would new call fcfae Bakshinayana, It s 
difficult to definitely ascertain the time when the commeiiee- 
t aiemfe of the year was changed from the vernal eqmnox to 
the winter solstice. But the change must have Ibeeik So- 
irodneed long before the verm! equtDox was in &e Exit- 
. * Taitt. Br. L 1. 2. 6 and L 5. 2, & * . 

t Sec TlioiBsoa f s Bfe^vacl ^ COL 


tik'is, and wlieii tin's change was made wttardyana, must 
have gradually come to denote the first half o the new 
year, i. <3;, the period- from the winter to the summer solstice, 
especially as the word itself was capable of being under- 
stood in the sense of '* turning towards the north- from' the 
southern-most point, " I am of opinion, however, that 
devaydn-a, and pitriy&na, or dkvaloka and pitrilok were the 
only terms used in the oldest times. It is a natural in- 
ference from the- fact th-afr the word uftwrdyana, as-such, 
does not occur- in- the- Rigveda.. The fact, that YisMvan 
was tlie central day of the yearly &atr>a, further shows that 
the sacrificial system was coeval with the division of the 
year- into the paths of Devas and" Pitris. After -a certain 
period the beginning of the year was changed to the winter 
solstice, and it was sometime after this change was made 
that the' words tiH-ardyuna- and $akshindy ana came to be 
used to denote the solstitial divisions of the year. But 
devaydna- and pitriydna could not be at once- divested of the 
ideas which had already become ^associated with them. 
Thus while- new feasts and sacrifices came to* be regulated 
according to- utt'ardyana and dakshindyana, devaydna- and 
fitriydna with all the associated ideas continued to- exist 
fey the sidfe of the new system, until they became either 
gradually assimilated with the new system or the priests 
reconciled the new and" the old systems by allowing option 
to individual's to follow whichever, they d'eemed "best. We 
most therefore take great care not to allow the idaa of 
uttardyaii(L % as we now understand it, to obscure our vision 
in, interpreting the- earjy Vedic traditions, and. that too 
much care can never be taken is. evident from, the fact that 
even so acute an astronomer as Bh&skar&ch&rya was at a 
to correctly understand, the tradition that the, Ut- 


tarayana was the Say of the De-vas. In his SiddMnta 
Shiromani tie raises the question how the Uttarayana, as it ; 

was generally understood in his day, could be the day of the ' 

Devas? He admits that the celestial beings om Meru at the j 

North Pole behold the si*n (during all the six months) 1 

when lie is in the northern hemisphere (vii. 9) and theso 
six months may therefore be properly called their day.* 
But tbe word uttardyana was then used to denote the 
period of six months from the winter to- the- summer sol- 
stice ; and Bhaskaracharya was unable to understand how ( 
such an Uttarayana could be called the day of the Devas by - l t 
tbe writers of the astronomical Sanhitas. If the sun is '' 
yisible to the Gods at Meru from the vernal equinox j'< 
to> the summer solstice, its passage back to the autumnal 
equinox lies through the same latitudes and in that passage, I 
i. e^ during the three months after the summer solstice, the 
sun must, says Bhaskaracharya, be visible to the Gods, 
But according to the Sanhita- writers the day of the Devas 
ended with the Uttarayana, that is, as Bh&skara understood 
the word, at the summer solstice. How is this conflict to 
be reconciled ? Bhaskaracharya could give no satisfactory 
solution of the difficulty, and asks his readers to reconcile- 
the conflicting statements an the supposition that the 
doctrine may be regarded as referring to "judicial astrology 

* In the Surya Siddhftnta xii. 67 it is said that "At Meru 
Gods behold the sun ? after but a single rising, during the half of 
liis revolution beginning with Aries ;" while in xiv, 9. the lit- J 

tarayana is said to commence "frona the sun's entrance inta Cap- 
ricorn." ' The author, however, has not noticed the tradition that 
the 'Uttarayana is the day of the Bevas and the apparent mconsis 1 - 
te.icy arising therefrom. Perhaps he understood the tradition in 
its true sense. 

II.] . SACRIFICE alias THE TEAR. 29 

and the fruits it. fore tells. 5 ' * Had Bhfiskanicharya however 
known that the word uf tardy ana was sometimes used for 
deiaydna, to denote the passage of the sun from the vernal 
to the autumnal equinox, I am suve^ he would uot have 
asked us to be satisfied with the lame explanation that the 
.doctrine of the Sanhita- writers need not be mathematically 
correct as it refers exclusively to judicial astrology. It is 
difficult to say whether the ancient Aryas ever lived so 
near the north pole as to be aware of the existence of a. day 
extending over at least two or three if not six rnontbs of the 
year. But the idea that the day of the Bevas commences 
when the sun passes to the north of the equator, appears to 
be an old one. In the Taittinya Brahmanaf iii. 9. 22. ]. 
we are told that the year is but a day of the Devas and 

* The original verses are as follows : 


Goladhyaya vii. 11-12, Bilpudevashustri's Ed. pp 304, 5, 

t 1^7 ^T ^rrt^r^TRf: 1 ^Tc^c^TC: I It is however extremely 
hazardous to base any theory upon this. Traditions like these have 
been cited as indicating the fact that the North Pole was inhabited 
in old days! Similar other traditions are said to indicate the 
existence of a pre-glacial period. Is it not more probable to 
suppose that when vt tardy ana and dcilcshinuyana, came to be first 
distinguished, they were respectively named 'day' and 'night' 
.with a qualifying word to mark their special nature ? The history 
of languages shews that when people come across new ideas 
they try to name them in old words. The Uttarayana and the 
pakshinAyana may hatv^ been thus conceived as Gods 1 day and 
night. See infra. Chap. V, 


even Herodotus (400 B. c-.) mentions a people who sleep 
during the six months of the year.* If Ihe tradition is, 
therefore,, as old as it is represented to be, it is impossible 
to reconcile it with- the later meaning of wttardyana as 
commencing from the winter solstice and this would then. 
furnish an additional ground to hold that in early times 

I the TJfctar&yana began with the* vermal equinox as stated in. 

I the Shatapatha Brahmana. 

I have stated above that when the commencement of the 
year was altered from fehe vernal equinox t the winter 
solstice^ uffiardyana either lost its older meaning or was 
rather used to denote the solstitial division of the year. 
But this is not the only consequence of that change. With 
the year the beginning of the annual satras was also 
gradually transferred to the winter solstice and the change 
was complete when the Taittiriya Sanhita was compiled. 
Tn fact had it not been for the passage in the Shatapatha 
Bi-ahmana it would have been impossible to produce any 
direct evidence of the older practice* When thje beginning 
of the satm was thus changed, the Yishiivan day must have 
gradually lost its primary meaning and come to denote 
simply the central day of the- yearly satra* 

The old practice was not however completely forgotten 
and for tte purpose of the Nakshatra-sacrifices the vernal 
equinox was still taken as. the starting point. Thus it is. that 
Garga tells us that " of all the Nakshatras the Krittikas 
are said to be tbe first for sacrificial purposes and Shr&- 
visbthl for (civil) enumeration.*' -f But even this disteao- 

* Quoted in l^airrien^s Origin and Progress of Astronomy, p. SI 
t Qaoted by Sonaakara on Tei. Jy. 5. 

: 1 


tion appears to have been eventually lost sight of by the 
latsr writers and all references to uttardyana were under- 
stood to be made solely to the six months from the winter 
to the summer solstice, an error from which even Bh&ska- 
racharya did not escape, though, he perceived the absurdity 
caused by it in some cases. At the present day we on the 
southern side of the Narmada begin the year at the vernal 
equinox for all civil purposes, but still all the religious 
ceremonies prescribed to be performed in the Uttarayana, 
are performed during the Uttarayana beginning with the 
winter solstice, a position quite the reverse of that described 
by Garg^a. When we at the present day have been 
thus using the system of a double year-beginning, we need 
not be surprised if the ancient Aryas, after shifting the 
commencement of the year to the winter solstice, managed 
to keep up the old and the new system together by assigning 
the different beginnings of the year to different purposes as 
indicated by Garga. It was the only alternative possible if 
nothing old was to be entirely given up. 




,' f 


Haksnatras in old Vedic times generally mean astcriams and not Zodiacal jsj- 

portions The present and the older position of the solstices In latex* ' 

works In Vedtnga Jyoti?ha An objection against its antiquity * s 

examined Passages in the Taittirtya Sanhit& and Br^hamana The &> 

KjittikAs head the Nakshatras Deva and Yania Nakshatras Their ^ 

real meaning Taittiifya Sanhita vii. 4 8. discussed Jaimiui's and 
Shabara's interpretation of the same~^Conclusions dedueible there- 
from- Winter solstice in ALagha 'Vernal equinox in the Kyittikas* 
Th agre of the Saahitli 2350 8. G* Bentley's arguments and " views 

WE HAVE seen that the ancient Aryas originally com- 
menced their year, which ^Yas luui-solar and siderial, with 
the vernal equinox, and that when the beginning \yas changed 
to the winter solstice both the reckonings were kept up, 
the one for sacrificial and the other for civil purposes. Let 
us now examine if there is any reliable evidence to show 
that "the Vedic priests made any corrections in the calendar 
when by the precession of the equinoxes the cycle of seasons 
gradually fell back. All our present calendars are prepared 
on the supposition that the vernal equinox still coincides 
with, the end of Revatt and our enumeration of the Naksha- 
tras begins with Ashvini, though the equinox has now 
receded about 13 from Eevati. It has been shown by Prof. 
Whitney (Surya Siddhanfca, viii.,, 9 note, p. 211 ) that theabove 
position of the vernal equinox may be assumed to be true 
at about 490 A.D. Taking this as the probable date of 
the introduction of the present system t we liave now to see 
if we can trace back the position of the veraaJ equinox 
amongst the fixed circle of stars. The question, so far as 


one antecedent stage is concerned, lias been thoroughly 
discussed by Colebrooke, Bentley, Max MiUler, Weber, 
Whitney, Biot, and other scholars ; and I shall therefore only 
summarise what they have said, noting the points where I 
differ from them. I do not propose to enter into any 
detailed mathematical calculations at this stage of the 
inquiry, for I am of opinion that until we have thoroughly 
examined and discussed all the passages in the Vedic works 
bearing on this question, and settled and arranged our facts, 
it is useless to go into minute numerical calculations. The 
Vedic observations could not again be such as need 
any minute or detailed arithmetical operations. I shall 
therefore adopt for the present the simplest possible method 
of calculation, a method which may be easily understood 
and followed by any one, who can wa.tch and observe the 
stars after the manner of the ancient priests. We shall 
assume that the zodiac was divided into 27 parts,not by 
compass but, by means of the leading stars, which. Prof* 
Max Miiller rightly calls the milestones of the heavens. 
The Vedic priest, who ascertained the motion of the sun 
by observing with his unaided eye the nearest visible star,* 
cannot be supposed to have followed a different method in 
making other celestial observations*; and, if so, we cannot 
assume that he was capable of recognizing and using for 
the purposes of observation any artificial divisions of the 
ecliptic on a mathematical principle, such as those which 
would result from the division of o60 of the zodiac into 2.7 
equal parts, each part thus extending over 13 20' of the 
ecliptic. Ofconrs^j auch an artificial method might bo easily 

* Taitt. Br. i. 5, 2, I, previously quoted. The passage is very 
important as it describes the method of making celestial obser- 
vations iu old times. . 



followed in later days, when the means of observation 
increased and the science of arithmetic was developed. 
But in the earliest days of civilization, it is more natural to 
suppose that the motions of the sun and the moon were deter- 
mined by observing which -of the known fixed stars was 
nearest to them. When we, therefore, find it stated in the 
Vedic works that the sun was in the Krittikas, it is more 
probable that the fixed asterism, and not the beginning of 
the artificial portion of the zodiac, was intended. I admit 
that the accuracy of such observations cannot be relied 
upon within two or three degrees, if not more. But we 
must take the facts as they are, especially when it is impos- 
sible to get anything more accurate from the ancient obser- 
vers of the heavens.* It will, I trust, however, be found 
that this inevitable want of accuracy in the old observations 
does not affect our conclusions to such an extent as to make 
them practically useless for chronological purposes. For 
instance, suppose that there is a mistake of 5 in observing 
the position of the sun with reference to a fixed star when 
the day and the night are of equal length. This would 
cause an error of not more than 5X 72=360 years in our 
calculations ; and in the absence of better means there is no 
reason to be dissatisfied even with such a result, especially 
when we are dealing with the remotest periods of antiquity. 
I shall, therefore, assume that references to the Nakshatras 
in the old Vedic works",* especially in cases where the mo- 
tions of other bodies are referred to them, are to the fixed 

* Similar observations have been recorded by Greek poets. 
Homer mentions ' the turns of the sun/ and Hesiod c the rising and 
the setting of the Pleiades at the beginnings of day and night.' 
The observations in the Vedic works may be supposed to have been 
made in a similar way. 


asterisms and not -to the zodiacal portions. I may also 
state here that as a change in the position of the vernal 
equinox necessarily causes a similar change in the position 
of the winter solstice, both the beginnings of the year, 
previously referred to, would require to be simultaneously 
altered. Whenever, therefore, we find a change in the 
position of the vernal equinox recorded in the early works, 
we must look for the evidence of a corresponding alteration 
in the position of the winter solstice, and the corroborative 
evidence so supplied will naturally add to the strength of 
our conclusions. This will, I hope, sufficiently explain the 
pi*ocedure I mean to follow in the investigation of the 
problem before us. I shall now proceed to examine the 
passages which place the vernal equinox in the Krittikas, 
beginning with the latest writer on the subject. 

It is now well-known that Varahatnihira, in whose time 
the vernal equinox coincided with the end of Eevatl and 
the summer solstice was in Punarvasu, distinctly refers in 
two places to the older position of the solstices recorded by 
writers who preceded him. " When the return of the sun 
took place from the middle of Ashlesha, " says he in his 
Pancha Siddhautika, " the tropic was then right. It now 
takes place from Punarvasu." * And, again, in the Brihat 
Sanhita iiL, 1 and 2, he mentions the same older position of 
both the solstitial points amd appeals to his readers to as- 
certain for themselves by actual observation which of the 
two positions of the solstices is the correct one, whether the 

* See Colebrooke's Essays, Vol. II., p. 387. The verse may now 
be found in Dr. Thaba-ut's edition of the work. It is as follows : 

- II 



older position of the solstices or that given by the writer.* 
It is clear, therefore, that in the days of Varahamihira, there 
existed works which placed the winter solstice in the begin- 
ning of (divisional) Dhanishtha and the summer solstice in 
the middle of Ashlesha. This statement of Varahamihira 
is fully corroborated by quotations from Gargat and Para- 
shara which we meet with in the works of the later com- 
mentators; and it appears that the system of commencing 
the year with the month of Magha, which corresponds with 
the above position of the solstices> was once actually in 
vogue. Amarasinha states that the seasons comprise two 
months each, beginning with Magha, aiid three such seasons 
make an ayana*$ The same arrangement of seasons is also 
mentioned in the medical works of Fhushruta and Vag- 
bhata. The account of the death of Bhlshma, related in 
the Mahabharata AnusMshaua-parva 167, further shows 
that the old warrior, who possessed the superhuman power 
of choosing his time of death, was waiting on his death-bed 
for the return of the sun towards the north from the winter 


: u 

t Giirg% quoted by Somakara on Yed. Jy. 5, says: 

: 11 

Bhattotpala on Brihat. San. iii. 1, quotes Garga as follows : 

: I , . 

J A mara i. 4. 18 

See rShushrab i. 6, and Vagbhata's Asht 


sthmm iii. 2; both of %hich are quoted further on it^ fchap. IV. 


solstice and that this auspicious event took place in the 
first half of the month of M&gha.* It is evident from this 
that the winter solstice must have coincided in those days 
with the beginning of Dhanishtha as described in the 
Vedanga Jyotisha and other works. 

There is thus sufficient independent evidence bo show that 
before the Hindus began to make their measurements from 
the vernal equinox in Revati there existed a system in. 
which the year commenced with the winter solstice in the 
month of Magha and the vernal equinox was in the last 
quarter of Bharani or the beginning of the Krittikas."f- We 
need not, therefore, have any doubts about the authenticity 
of a work which describes this older system and gives rales 
of preparing a calendar accordingly. Now this is what the 
Vedanga Jyotisha has done. It is a small treatise on the 
Vedic calendar, and though some of its verses still remain. 
unintelligible, yet we now know enough of the work to 
ascertain the nature of the calculations given therein. It 
once supposed that the treatise mentions the Rdshis, 

* Mah. Ann. 167, 26 and 28: 

Lele, Modak, fcetkar and other Hindu astronomers have recently 
tried to determine the date of the Mahabharata war from such 
references, and they hold that the vernal equinox was then in the 

t Prof. MaxMuller has pointed out that in the Atharva Veda 
i. 10. 7 and in the Yajnavalkya Smriti I. 267, the Krittikas occupy 
their early position, while the Vishnu Parana actually places the 
vernal equinox in theKpfctikas* See Pref. to Rig,, Vol. IV,, p. xxxi. 




but a farther study of the work has shown that though the 
word Rdshi occurs in some of its verses, it is there used in 
a totally different sense. This work gives the following 
positions of the solstices and the equinoxes: * 

1. The winter solstice in the beginning of Shravishtha, 

(divisional) ; 

2. The vernal equinox in 10 of Bharani; 

3. The summer solstice in the middle of Ashlesha, 

4. The autumnal equinox in 3 20' of Vishakha. 

The first year of the cycle commenced with the winter 
solstice* when the sun and the tnoon were together at the 
beginning of Dhanishtha and the Uttarayana also began at 
the same time. There is very little else in the Vedanga 
Jyotisha that may help us in our present inquiry except the 
fact that the enumeration of the deities presiding over the 
various Nakshatras begins with Agni, the presiding deity 
of the Krittikasf. From these data astronomers have cal- 
culated that the solstitial colure occupied the position above 
mentioned between 1269 B. C. to 1181 B. C., according as 
we take the mean rate of the precession of the equinoxes 
50* or 48*-6 a year.J 

Some scholars, however, have boldly raised the question, 
what authority is there to hold that the position of the 
solstitial colure was recorded in the Vedanga Jyotisha from. 

* See Ved. Jy, Verse 5. 


t.Cf. Ved. Jy. Verse 25, ^ft : spHTT^r. ^Tfr &c., 
J See the late KrishnasMstri Godbole's Essay on the Antiquity 
of the Vedas., p. 18 ; also Pref. to Rig., Vol. IV., p. xxviiL 


actual observation ? It is conceded that the position of the 
solstitial colure might have been incorporated in the Jyot- 
isha from real traditional information, but it is at the same 
time contended that the language of the treatise and the 
methods given therein create doubts about the antiquity 
claimed for the work on the strength of the position of the 
solstitial points given therein. "I feel bound to remark/' 
says Prof. Max Miiller, " that unless there was internal 
evidence that the Vedic hymns reached back to that remote 
antiquity this passage in the Jyotisha would by itself carry 
no weight whatever*/ 5 The existence of the different versions 
of the Vedanga Jyotisha and the obscurity into which some 
of its verses are still shrouded render it rather difficult to 
meet the above objection, especially as it is a side attack on 
the antiquity of the work with an admission that the posi- 
tion oE the colure might have been recorded in the work 
from real traditions current in the time of its author. It 
is, however, needless to answer this objection, inasmuch as 
there is ample confirmatory evidence in the Vedic works 
themselves which not only bears out the statement in the 
Vedanga Jyotisha, but takes us back into still remoter 

There are many passages in the Taittirlya Sanhita, the 
Taittiriya Brahmana and other works where the Krittikas 
occupy the first place in the list of the Nakshatras.t In 
the Taittiriya Brahmana (i. 1, 2, 1) it is distinctly stated 

* See Pref. to Rig., Vol. IV., p. xxv. The mention of ^drr for 3T^- 
2pf, first in the list of symbolic representations of the Nakshatras 
in verse 14, lends some support to these doubts. 

t These together with the list, will he found in Pref. to Rig., 
Vol IV., p. xxxiv. Of. Taitt. San. iv., 4. 10; Taitl Br. iii. 1. 1. 6 and 
i.5. 1. 2. 




"one should consecrate the ( sacred ) fire in the Krittikas; ^ 

......the Krittikas are the mouth of the Nakshatras."* ^ 

This shows that the first place given to the Krittikas in the ^ 

list of the Nakshatras is not accidental and that we must at ^ 

least suppose that the Krittikas were the *' mouth of the ^ 

Nakshatras/ 5 in the same way as Vasanta or spring was the ^ 

"mouth of the seasons'^ or the Phalguni full moon the J'j 

" mouth of the year.":}' The phrase is the same in all places ^ 

and naturally enough ifc must be similarly interpreted. But h * 

granting that the Krittikas were the mouth of the Naksha- y 

tras in the sense that their list always commenced with them, ^ 

it may be' asked what position we are to assign to the I 

Krittikas in the course of the year. There were, as I have f 

previously shown, two beginnings of the year, the winter F 

solstice and the vernal equinox; which of these two cor- | 

responded with the Krittikas ? Or, are they, to be supposed j r 

to have coincided with a point altogether different from | , 

these two ? A little consideration will show that it is not f 

difficult to answer these questions satisfactorily. The | 
present distance between the Krittikas and the summer 
solstice is more then 30, and if they ever coincided with 
the summer solstice it must have been long ago in .the pre- 
sent cycle of the precession of the equinoxes. We cannot 
therefore interpret the above passage so as to place the 
summer solstice in the Krittikas, unless we are prepared to 
take back the composition of the Taittiriya Sanhita to about 
22,000 B, 0.5 and further suppose that all evidence of the 
intermediate astronomical observations is entirely lost, and 
the same thing may be said against placing the Krittik&s 

t Tail*. Br.11.2. 

$ Taitt. San. vii. 4. 8 cpdted infra. 

f f 'i 


in the autumnal equinox.* Both the suggestions In my 
opinion are too extravagant to deserve any consideration. 
Nor can we assign the beginning of the Nakshatras to any 
random point in the ecliptic. There thus remain two pos- 
, sible explanations: viz., that the Krittikas coincided either 
with the winter solstice, or with the vernal equinox, 
Now, considering the fact that the vernal equinox is placed 
in the last quarter of Bharani in the Vedauga Jyotisha it is 
more natural to presume that the vernal equinox coincided / 
with the Krittikas at the time when the Taittirlya Sanhita / 
was compiled. But we need not depend upon probabilities 
like these, when there are other passages in the Taittiriya 
Sanhita and Brahmana which, serve to clearly define the 
position of the Krittikas in those days. 

In the Taittiriya Brahmana (i. 5, 2, 7) it is stated that 
" the Nakshatras are the houses of gods , . . . the Nakshatras 
of the Devas begin with the Krittikas and end with Vishtakha, 
whereas the Nakshatras of Yama begin with the Anuradhas 
and end with the Apa-Bharanis. J; t Prof. Max Mliller 
appears to think that the latter group is called the Naksha- 
tras of Yama because Yama presides over the last of them.J 
But the explanation appears to me to be quite unsatis- 
factory; for on the same principle the first group should 
have been called the Nakshatras of Indr&gni, the presiding 
deities of VishakM, the last in that group. I am, there- 

* A similar mistake is committed by the late Krishna Shastri 
Godbole, in his essay on the antiquity of the Veclas, where he 
supposes Mrigashiras f?o be in the autumnal equinox, p. 20, 21. 

t %^j?r I wsrrfa i - - v- i frftrarn *nrt I ft^rr^ ^rrt i 
i arfncrwr- TO* larwt^ff^tR ^rf^r ^R$n*Tf*r | 
nvr ffaprV ?t:4rW i *m 

Pref. to Rig. Vol. IV. p. xxxi 


fore, disposed to think that tlie principle of division in this 
case Is the same as that followed in the case o the Devayana 
and the Pitriyana discussed before. We have the express 
authority of the Shatapatha Brahmana stating that the sun 
was to be considered as moving amongst and protecting the 
Devas, when he turned to the north, in the three seasons of 
spring, summer and rains. In other words the hemisphere 
to the north of the equator was supposed to be consecrated 
to the Devas and the southern one to the Pitris. Now, the 
sun moved amongst the Devas when he was in the northern 
hemisphere. The Devas, therefore, must have their abode 
in that hemisphere, and as the Nakshatras are said to be 
the houses of the Devas, all the Nakshatras in the northern 
hemisphere, from the vernal to the autumnal equinox, would 
naturally be called the Nakshatras of the Devas. Now the 
southern hemisphere was assigned to the Pitris; but I have 
already quoted a passage from the Rigveda which states 
that it was the path of the god of death. In Rig. x. 14, 1, 
Yama is spoken of as the king of Pitris, and in verse 7 o 
the same hymn the deceased is told to go to the pitri-loka, 
where he would meet the god Yama. In the Vajasaneyi 
Sanhita 19, 45, salutation is made to the world of Pitris in 
the kingdom of Yama. There are many other passages of 
similar import in the Sanhitas,* and from all these it would 
be quite clear that the Pitriyana or Pitri-loka was also called 
the kingdom of Yama. The Nakshatras in the southern 
hemisphere, therefore, came to be designated as the Nak- 
shatras of Yama in opposition to the Nakshatras of th^ 
Devas, thus dividing the whole circle of stars in two equal 
groups. This also explains why Yama is made to preside 

* Cf.Taitt. San. vik.3, 14. 3jiR 
[ I Also see A that. Ved, xviii, 4. 


over the Apa-Bharanis. It was at tlis Apa-Bharanis that 
tlie zodiac was divided, the Krittikas going over into the 
Devas* and the Apa-Bharanis turning* down into- the Yama's 
portion of the celestial hemisphere.* The Taittiriya Brah- 
mana further states that the Nakshatras of the De^as move 
towards the south while the Nakshatras of Yama move 
towards the north. The words dakskina (south) and wttara 
(north) are in the instrumental case, and doubts have been, 
entertained as to their exact meaning. But if we accept 
the statement in the Shatapatha Brahman a about the abode of 
the Devas, no other meaning is possible 'except that the 
Nakshatras of the Devas were counted from the vernal to 
the autumnal equinox:, that is, to the point where the south 
(southern hemisphere) begins, and conversely in the ease of 
the Nakshatras. of Yama. I may here mention that the 
movements of both the groups are described in the Brahmana 
in the present tense (pari-yanti), and that we may, therefore, 
suppose the-m to be recorded from actual observation. If 
this explanation of the division of the Nakshatras. into the 
Nakshatras of gods and those of Yama is correct and I 
think it is it at once fixes the position of the Krittikas at 
the beginning of the Devay&na or the vernal equinox at the 
time when these Vedic works were compiled,. 

There is another and still more important passage in the* 
Taittiriya Sanhita which supplies further confirmatory evi- 
dence on the same point. In the Taittiriya Sanhita vii. 4, &, 

* May not Apa-Bharanis have heen so named from this circum- 
stance ? Bharam appears to be an older name, changed afterwards 
into Apa-Bharani, in the same man-ner, Mula into Mula-barham, and 
Jyeshtha into Jyeshfrhaghni. Perhaps the description of Apa-, 
rani in Taitt, Br. 1 5, 1, may be so understood. It says : 



we have a*dlscussion as to the time best suited for the com- 
mencement of the Satras like the gavam-ayana which last for 
one whole year and as the passage is important in various 
ways I shall give it here in the original: 

reft ^^rF%<re*T rqrrr^ uf^r 



In the T&ndya Brahmana (v. 9) we have the same pas-* 
sage with a few additions and alterations, and as this has 
been quoted by the commentators I shall give it here for 

M ^ n 



II ^ H 

The third sutra in the above gives an additional reason 
for rejecting the Ekaahtaka ; while in the fourth, sutra vich- 
hinnam is substituted for vyastam of the Taittiriya Sanbita. 
^Jiother important cnangeis, that the word Phalguni~'p6rna- 
mdsa is paraphrased by PMlguni in the 8th #&tra, 


clearly showing that the former was then understood to 
mean the fall-moon night. Both the passages are similar 
in other respects. 

Fortunately for us Sayanacharya is not our only guide 
in the interpretation of these important passages. It is 
probably the only passage (the two passages being similar 
I treat them as one) in the Vedas where the commencement 
of the annual satra is given and from the ritualistic point 
of view it has formed the subject of a learned discussion? 
amongst the Mimansakas. Jaimini in his Mimans&-darshana 
Chap. vi. Sect. 5 has devoted an Adhikarana (10th) to the 
interpretation of this passage, and the subject has been 
thoroughly discussed by Shabara, Rumania, Parthasarathi, 
Khandadeva and other writers on Mimansa. We have thus 
a continuous tradition about the meaning of this passage 
current amongst the Indian divines a tradition based not 
upon mere authority, but on the logically solid; rules of 
exegetics propounded in the work of Jaimini. I shall first 
give a literal translation of the passage from the Taittirfya- 
ISanhita and then discuss its interpretation as bearing on, the 
present question. 

"Those who are about to consecrate themselves for the 
year (sacrifice) should do so on the Ekdshtaka (day). The 
EklshtaM is the wife of the year ; and he [i. e., the year] 
lives in her f> e., the Ekashtaka] for that night. (Therefore 
they) practically sacrifice (by) beginning the year.* Those 

* The Taadya Brihmana, (S&tra 3 in the above passage) ' 
~a third reason thus: -"They go to avabhntha [i.e., the final bath] 
not delighted with water/' Shabara and other commentators 0*1 
Jairaini have noticed this additional ground for raeetin* the 
^EkashtaM. , D 



that sacrifice on the Ekashtaka, sacrifice to the distressed 
( period ) o the year. It is the season .( dual) whose name 
comes last. Those, that sacrifice on the Ekashtaka, sacrifice 
to the reversed* ( period ) of the year. It is the season 
(dual) whose name conies last. They should consecrate 
themselves for the sacrifice on the Phalguni full moon. 
The Phalgnni full- moon is the mouth of the year. They 
sacrifice (by) beginning the year from the very mouth. It 
has only one fault, viz., that the Vishuvan [i. e>, the equator 
or the central day] falls in the .rains. They should conse,- 
crate themselves for the sacrifice on the Chitrci, full-moon. 
The Chi bra full-moon is the mouth of the year. They 
sacrifice (by) beginning the year from the very mouth. It 
has no fault whatsoever. They should consecrate them- 
selves for the sacrifice four days before the full-moon. 
Their Kraya [i. e>, bhe purchase of soma.] falls on the Ekash~ 
taka. Thereby they do not render the Ekashtaka void 
[i.e., of no consequence]. Their Sutya [i.e., the extraction 
of soma juice] falls in the first [i.e., the bright] half (of the 
month). Their months [i'.e., the monthly sacrifices] fall in 
the first half. They rise [i.e., finish their sacrifice] in the 
first half. On their rising, herbs and plants rise after them. 
After them rises the good fame that these sacrificers have 
prospered. Thereon all prosper." 

Here in the beginning we are told that the Ekashtaka is 
the day to commence the Satra, which lasts for one year. But 
the word Ekashtaka is used to denote the eighth day of the 
latter (dark) half of the four months of Hemantaand Shishira 
seasons, f and sometimes it means the eighth day of the 

* According to the TandyaBrahmana"hroken' J or "destroyed." 
t Cf. Ashv. Gri. Sutra iu 4/1; 




flark half of each of the twelve months of the year.* 
The statement iu the following sentences that this lilkashtaka 
is liable to the objection of occurring in the cold or the 
last season does, however, at once narrow the field of 
our choice. It must be further borne in mind that the 
Ek&shtaka, here spoken of, is the wife of the year, and is 
contrasted with the Phalguni and Chitra full-moons; while 
tradition in the time of Jaimiui and Apastamba interpreted 
it to mean the 8th day of the dark half of Magha. All 
writers on Mimansa therefore take this Ekashtaka to mean 
the 8th day of the dark half of Magiia. As the Ekashtaka 
is the wife of the yearf and as the god of the year is said 
to reside with her on that night, those that commence their 
sacrifice on the Ekashtaka may practically be supposed to 
commence it at the beginning of the year which resides 
there. In other words the Ekashtaka is thus a constructive 
beginning of the year, and therefore the yearly sacrifice 
may be commenced ou that day* Bat the passage now 
proceeds to point out the objections to the commencing 
of the sacrifice on the Ekashtaka day. The 8th day of 
JSagha falls during the distressed period of the year, that 
is* according to Shabara and other commentators the period 

Of. Taitdya Br&hmaua x. 3, 11. 

his commentary on fan. Br. v. 9, observes that Ekasktakfi 
is there wsed inks secondary sense and quotes Apastamba Grihya 
Sutra (iiL 21, 10) thus: 2ff TT^T qfK*TR*CT ^TKSTSIS-^r rf^fST'fr 

^ t Thus both Jaiminiaud Apastamba 

considered Ekasfetakato mean the 8th day of the dark half of 

f Shabar on Jairnini vi. 5, 35, quotes Atharva Veda iii. 10, 2 i 
.and Sayana in his Comra. on Taitt. San. vii. 4, 8, cites, Atharra 
"VednitL 10, i^, and Taitt. Sun. iv. 3, 11, 3. But these texts simply 
state ihat the Ekfchgaki is. the wife of tiie^year... . .without'- 
the Bkashtaka. 


when people are distressed by cold.* The word in the text is 
Aria which literally means 'distressed,' but Sayana takes it to 
denote the end or destruction of the year, implying thereby 
that the old year is then brought to an end and that the 
consecration for the yearly satra, which must be made before 
the beginning of the new year, or, in other words, not after 
the previous year is ended or destroyed, cannot be made at 
this time. Though Shabara and Sayana thus give different 
interpretations of drta, practically both agree in holding 
that in those days the old year ended before the eighth 
day of the dark half of Magh ; for Shabara distinctly states 
that the word '' reversed " used further on means " reversed 
on account of the change of ayana""\ Sacrificing during. 
the distressed period of the year is thus the first objection 
to commencing the satra on the Ekashtak day. The 
second objection is that it is the last season, that is, though 
you may be said to sacrifice to the constructive beginning 
of the year, yet as far as the seasons are concerned you 
sacrifice in the last of them. The word for season, ritu, has 
been used in the text in the dual number and it might be 
urged that it denotes two seasons. A reference to the 
Taittiriya Sanhita iv. 4-11, 1. will, however, show that the 
word ritu is there used in dual, probably because each 
season comprises two months, just as "scissors" is 

Shabara on Jaino. vi. 5. 37, Sayana 

I Shabara on Jaimini vi. 5. 37. Sayana 

in his comm. on Taitt. San. says 

srr^ fffr^rr 

Upon this passage the author of 

Kaia-Madhava observes ffT^nf ^^^"^Tr^rPT^^ I Cal. Ed., -p. 



; used in plural in English. A similar passage also occurs 

in the V&jasaneyi Sanhita (13-25) and Mahidhara while 
commenting on it expressly states that the dual there has the 
: meaning of the singular number.* The " last named seasons 

(dual)" therefore simply means " the last season/' It must 
; be here mentioned that according to the passage in the 

; Tanclya Brahmana, which Shabara appears to quote, the 

) firsfc objection is thus stated : " not delighted with water 

' they go to avabhritha [i e., the final bath] ." This is but an 

; amplification of the objection on the ground of the "last 

i season ** and Khandadeva expressly says that water is then 

{ nadeiigbtfal * tf on accomnfe of cold." The Tandya Brahmana 

* -not omil the objection- of the "last season;" but 

simply expands and illustrates the same by referring to the 
nataral dislike for a cold bath in that season. We may., there- 
fore, regard this objection, more as explaining the first than 
as an additional one. We now come to the third objection, 
viss.y those that commence the sacrifice on the Ekashtaka day 
sacrifice to the reversed period of the year. 'Keversed/ 
L the original, is said by Shabara to indicate the change 
caused by the turning away of the sun from the 
itiee/f- and Sayana seems to understand it in. the 
'ttffb Thus although those that commence the satra 
4Mt the Sib day o the dark half of MagM may be sup- 
to do s0 tactically at the beginning of the year, the 
; of Ite &:fehtsk& ? yet ikbe procedure is triply objec- 
ae tty sacrifice in the cold season, in 
last of (wbeiD w&fer is tmdrfigttf al) and wliexx 

i or *by the turning away of tite istis 

from &Q 

t Bee second note on the last page. 

' '''M^ ' i ' ' ' ^ ' v * ^ ' 

^L'uLyVi^Siii&yj 1 ! , *>' * ; :"- r / . ''* > r 


To get over this threefold objection an alternative 
next proposed. The Phalguni full-moon day was 
to be the first day of the year. If you commence 
sacrifice on that day, you avoid the three objections pre- 
viously noted and still secure your object of sacrificing at 
the beginning or the mouth of the year. But even 
course is not faultless -, because if you commence on the 
Phalguni full-moon the middle or the central day of 
falls in the rainy season, which again is not a desirable 
The first twelve days of a satra are taken up in 
cration and twelve more in upasacte after which the 
satra sacrifices commence. So the middle day of tlic 
falls after six months and twenty-four days from the Phal- 
guni full-moon, that is, on the ninth oi the bright, of 
the month of Ashvina."^ Now if we suppose 
solstice or the beginning of the cold season to on the 
Magha full-moon, the summer solstice, or the of tbi* 
summer and the beginning of the rainy sess0a y fell 
a little after the full-moon in Shravana. The 
drapada and Ashvina therefore represented the 
in those days, and the occurrence of the VishAfia in 
or the rainy season was not believed to be auspicious. Aa 
the next alternative it is, therefore, suggested the 
consecration should take place on the Chitrfr full-moon, and 
this, course is said to be open to no objection 

But even this is given up for a still better time, and it is 
finally stated that persons desirous of consecrating 
selves for the safra, should do so "four days ike fmU 

moon." The full-moon here mentioned is not, 
specifically defined,, ..and consequently it forms the 

* This, in suhstance, is S&ywja's explanation in hit 
on this passage. 


4fe T.ffi ORION. [CHAPTER 

of one of the Jaimini's Adhikaranas.* As no specific full- 
moon is mentioned it may mean either any full-moon-clay", < 
or the Ghitra full-moon which is mentioned next before in 
"the same passage, or it may refer to the Magha full-toooii 
as the Ekashtaka is mentioned immediately afterwards in ( 
connection with it. Jaimini decides that it is the full-moon 
in the month of Magha, for it is stated immediately after 
that those who commence the sacrifice on this full-moon will 
purchase their Soma on the Ekashtaka. This Ekashtaka: can , 
evidently be no other than the one mentioned in the begin- 
nisg of the passage, and the object of the arrangement last I 
suggested is to utilise somehow or other the important day f 
" the Skashtaka, which was at first recommended for the ,f 
commencement of the sacrifice itself, bat which had to be ,j? v 

5; 2. ^Kdift*qr5; % *TTsfr l^rmg^ft: ; 4 

6. frWTO^ ^trHHff^f ?: 

f 8. 3ittr ^ ^nf 1*5*11 M. Sajana ia his Jaimim-nyaya-mala 
Yist&ra aadin his comnu on the Taitt. San. fully adopts this view. 
But ia Ms eomai.,011- tbe Tandya Brahraana, v. & 12 (Cal. Ed.) he 
is represented as saying that the fail-mooa last mentioned refers to 
tlie pbaitei ! S^ ^ ^ther the seribe, the pfmter^ or the 
publisher, has here oferitomfy committed an error* * i 

given up on account of the three-fold objection stated above. n 

The full-moon must, therefore, be the one next preceding this I 1 

Efcashtafca. Again the full-moon day is said to be such that ''* 

when the sacrifice is finished the herbs and the plants spring ' 

up, which, as remarked by Shabara, can happen only in the ^ 

Vasanta season. !| 

*Eb sum up; the last mentioned fall-moon, though not 2 

specifically defined, must be prior to the Vasanta season I 

'* Jalmini vi 5> 30*37. Jaimini's Sutras which I have here tried /, 
to transkte arf explain are as follows : 1. 


and also the next previous to the Ekashtaka, which is the 
wife of the year and which falls in the cold season, in the 
last season, i. e., Shishira [or when water is not delightful! 
and after the sun has passed through the winter solstice. 
It must also be remembered that the Phalguni and the 
.Chitra full-moon are to be excluded, Jaimini, therefore, 
concludes that this full-moon cannot be any other than the 
one falling in the month of Magha, and his conclusion 
been adopted by all the Mimansakas. We can now under- 
stand why Laugakshi, quoted by Somakara, stales 
"they sacrifice to the year four days before the full-moon 
in Magha. "* 

If Jaimini' s interpretation of this passage is correct, we 
may, so far as our present inquiry is concerned, deduce 
the following conclusions from it : ( 1 ) That in the days of 
the Taittiriya Sanhitft the winter solstice occurred before 
the eighth day of the dark half of Magha, which again wast 
a month of the cold season. Whether the solstitial day fell 
on the Magha full-moon is not so certain, though it mmj be 
taken as fairly implied. For the Ekashtakl was abandoned 
because it occurred in the "reversed" period of the year, 
and it is quite natural to suppose that the priests in 
choosing a second day would try to remove as many of 
the objections to the Ekashtaka as they could. In other 
words, they would not select a day In the "reversed " period 
of the year, nor one in the last season. The fact that a day 
before the fall-moon in Magha was selected is, therefore, a 
clear indication of the solstice occurring on that day, white 
their anxiety to utilise the EkashtaM fully accounts for the 
selection of the fourth in preference to any other day before 
the full-moon. I may also remark that throughout the 



whole passage the intention of sacrificing at the beginning 1 

(real, constructive, or traditional) of the year is quite clear. > 

The full-moon in Magha must, therefore, have been one of ^ 

such beginnings, (2) That the year then commenced with *'/ 

the winter solstice. (3) That as there cannot be three real 1 

beginnings of the year at an interval of one month each, the ,V 

passage must be understood as recording a tradition aboub r 

the Chitra full-moon and the Phalguui full-moon being onco |f 
considered as the first days of the year. (4) That Vishnvun 
had lost its primary meaning and that it fell in the rainy 
season if the sacrifice was commenced on the Phalguni 

The passage thus supplies not only confirmatory, but 
direct evidence of the coincidence of the Krittikas with the 
vernal equinox in the days of the Taittiriya Sanhitu. For, 
if the winter solstice fell on the full-mocn day in Maghd, 
then the summer solstice, where the moon must then bo . 
must coincide with the asterism of Magha, and counting 
seven Nakshatras backwards We get the vernal equinox in 
the Krittikas. Independently of the Vedanga Jyotisha we 
ttns have four different statements in the Taittiriya Sanhitft 
and Briliinana clearly showing that the vernal equinox was 
thenia fete Krittikas": firstly, the lists of tbe Nakshatras 
and their 'presiding deities, given in the Taittiriya SanhitS. 
and Brttmana, all beginning with the Krittikas ; secondly, 
an express statement in the Taittiriya Brahraana that the 
Krittikas are the mouth of the Nakshatras ; thirdly? a 
statement that the Krittikas are the first of the Ddva itilc- 
shatras, thai is, as I have shown before, the Nakehakw in 
the northern hemisphere abov& the vernal eqtuttog; j and 
fourthly, the passage in the Taittiriya 'Sajabitfc above dis- 
cussed, which exprmif abates that th Iriatar r solstice fell 


in the month of MagM. The vernal equinox is referred to 
the Krittikas directly or indirectly in all these passages 
and I do not think that any more confirmatory evidence 
from, the Vedic works is required to establish the proposition \ 
thatuihe Krittikas coincided with the vernal, equinox, when ' 
the Taittiriya Sanhitft was compiled. I As an additional 
proof I may, however, mention the fact, that Pitris 
are said to be the presiding deities of Magh& in the 
Taittiriya Sanhita iv. 4. 10. ]. With the Krittikas in the 
vernal equinox Magha iS at the summer solstice and as 
the Dakshinayana or the ay ana of the Pitris commenced- 
at this point, the asterism which happened to be there at 
that time was naturally assigned to the Pitris. The position 
of all the other, cardinal points of the ecliptic can be thus 
shown to be consistent with the position of the vernal 
equinox in the Krittikas. 

Supposing the Krittikas to denote the asterism of that 
rxarne this gives us, according bo Prof. Whitney's* calcula- 
tion, 2350 B. C. as the probable time for the compilation 
of the TaittirJya Sanhita. Some scholars unwilling to carry 
t}ie antiquity of the work to such a rempte period, have 
urged, without assigning any special reason, that by Krit- 
tikas we must here understand the beginning of the zodiacal 
portion of that name. Now as the position, of th^ asterism. 
of the Krittikas in its zodiacal portion is 10 50' from tho 
beginning^ these scholars would place the vernal equinox 
about 11 behind the asterism of the Krittikas and thus 
reduce the antiquity of the Sanhitfr nearly by 1 1 X 72=792 

* See Surya Siddhftnta Add. notes, p. 323, 
t This is the position given in the Surya Siddhanta viii. 2-0. 
See the table prepared by Prof. Whitney in his notes to thia 


years or to about 1426 B. 0.* I have briefly stated before 

my reasons for discarding this supposition and holding that 

the names of the Nakshatras in the early Vedic days must *" 

be taken to denote the asterisms known by such names. If 

Indian priests are to be supposed incapable of making any 

accurate observations of solstitial points in J 200 B. C.,f it 

is to my mind utterly inconsistent and illogical to hold that 

the forefathers of these priests, when they assigned the 

vernal equinox to the Krittikas, understood the word to 

mean not the asterism but the imaginary beginning of the 

zodiacal portion of that name. / I cannot also understand 

why scholars should hesitate to assign the Vedic works to 

the same period of antiquity which they allow to the Chinese 

and the Egyptians.f Bufc it is needless here to euter into 

this, controversy. For if I once succeed in showing, as I 

hope to do. that there is sufficient internal evidence in the 

Vedic literature itself of a still remoter antiquity, all theories, 

conjectures, and guesses, which have the effect of unduly 

reducing the antiquity of the Vedic works and also of 

throwing discredit upon the claims of the Indians to the 

c-riginof the Nakshatra system, will require no refutation.) 

Bentley, however, takes his stand on a different ground. 

* This is Bentley's date about which see 'infra. 

t See Prel. lo Rig., vol IV., p. xxix. 

| M. Biot allows it> the case of the Chinese and considers 
* Rt A ! h ! . Hindus k' )rrow ed ^e Nakshafcra system from them. 
AJbtrteJ, in bi3 chronology of ancient nations, <fcc., observes that 
other nations begin their asterisms with the Pleiades. He further 
tfalfe that he has found in some books of Hermes that the vernal 
qimox ompctdrt'tftb the.rising of the Pleiades, but, sap he, "God 
knows best what tkey is* ' m ' " 




He suggests that the word VishdkM, like Vidala,* may 
mean "possessed of two branches/' and that these two 
branches may have been caused by the equinoctial colure 
bisecting the zodiacal portion of the Vishakhas. Now the 
equinoctial colure passing through the beginning of the 
divisional Krittikas naturally bisects the zodiacal portion of 
VishakM. Bentley, therefore, concludes, without any more 
proof than this etymological conjecture, that this was the 
position of the colure when Vishakha received its name. 
This is no doubt an ingenious hypothesis. But there is 
not only no evidence in the Yedic works to support such 
etymological speculation, but it may be easily shown to be 
inconsistent with the position of the winter solstice in 
the days of the Taittiriya Sanhita'. 

I have already stated that from the passage of the Tait- 
tiriya/ Sanhita just quoted we may fairly infer that the 
winter solstice occurred in those days on the full moon in 
M&gha. According to the Vedanga Jyotisha it fell a fort- 
night earlier, that is, on the first day of the bright half o 
Mgha. It is roughly estimated that the equinox must 
recede about two divisional Nakshatras, i.e. 26 40', to make 
the seasons fall back by one month. Between the times of 
the Taittiriya Sanhita and the Vedanga Jyotisha the equinox, 
must accordingly recede 13 20' or nearly 14. Now the 
position of the equinox as given in the Vedanga Jyotisha is 
10 of Bharani From this to the beginning of the divi- 
sional Krittikas, the distance is only 8 20', while if we mea- 
sure it from the asterism of Krittika it is 3 20'+ 1 50' = 1 4 
10 r . Therefore during the period that lapsed between the 

* This example has been added by Prof. Max Miiller. See Pref. 
to Rig., Vol. IV., p. xxx. See also Bentley 'a Historical view of 
Hindu Astronomy, p, 2. 


Taittiriya' Sanhita and the V-edaaaga Jyotisha the equinox^ 
according to JBentley, receded only 3 20' ; while if we 
understand the Kritfcik&s to denote the asterism of that 
name, it .gives as a precession of 14 10'. Now as the winter 
solstice fell a fortnight later in the days of the Sanhita we 
must accept the latter precession of 14, which alone 
corresponds with that interval of time (i. e. a fortnight) and 
assume that the vernal equinox then coincided with the 
asterism of Krittika, a conclusion the probability of which 
has already bean -established on other grounds. Bep.tley*$ 
speculation must, therefore, be rejected, unless we ar 
prepared to allow his guess about the primary meaning of 
Visbakba to prevail against reasonable conclusions based 
upon a passage from the Taittiriya Sanhitft. , 

7-V ' ' 

But even admitting Bentley's speculation about the 
meaning of Vishakha, we may fairly question the soundness 
of the conclusion drawn therefrom. For what ground 
is there for holding that the two divisions of Vishakha must 
be mathematically equal in every respect ? The word dala 
in v-idala may be so understood ; but dala and shdkkd are 
not similar in this respect. Bentley's error., therefore, con- 
sists not in supposing that tha colure may have cut the 
divisional Vishakhas, but in inferring therefrom that it 
must have bisected it. The whole ecliptic was divided 
into 27 Nakshatras, and 13 could only be comprised in 
each hemisphere. YishakM, the 14th ISTakshatra from 
the Krittik&s, may have been thus considered, by simply 
counting the number of the Nakshatras, as lying partly 
in the region of the Devas and partly in that of the Pitris.* 

to satisfy a tamely etymological speculation 
evidence whatsoever* Speaking more 


For though we might hold that the Vedi observers were 
not provided with means to fix imaginary points in the 
heavens and ta refer to these points the motions oi the 
heavenly bodies as- astronomers do at present^ yet it does 
not imply that they were unaware oi the approximate dis- 
tances between, the various asterisms selected by them* In? 
other words,, they might be supposed to have, roughly 
known the distances between the stars, though for obvious- 
reasons they could not bat refer the motions of the heavenly 
bodies only to the fixed stars.. Thus understood^ Bentley '& 
conjecture about the primary meaning of Vishakhi does 
not necessarily imply that, the equinoctial colure bisected 
the divisional VishakMs in those days * r and when the 
conjecture itself does not thus support his theory about 
the position of the coliire, I do not think we shall be justi- 
fied in accepting it, especially when it is- shown, that it is 
also objectionable on ather grounds.. I am, thei?efoBe^ dis- 
posed to fix the date of the Taittiriya Sanhita-at 2350 B. C^ 
and not 1426 B. O 4 . as Bentley has done. 

accurately if the vernal equinox, coincided with the- asterism of the 
Krittik&s, the equinoctial colure falls out of the divisional Vishakas 
by 4, but ife is nearly 6 behind the asterism of Aniiradha. Of these 
two asterisms Yishaklui would therefore be nearer to the colure. 
But we might as well ask what ground there is for holding that the 
Nakshatra division* of the Zodiac, at the time when the vernal 
equinox was in the Krittikas (supposing sueh divisions to have then 
existed), were the same as those which we-now use and which- com- 
mence with Revati. Bendey appears to have altogether overlooked 
this objection. I have already stated my view regarding the exist- 
ence of the divisional Kakshatras in old times,, and I would reject 
Bentley 7 s etymological speculation oa the mere ground that it 
requires us to assume the existence of such diriskmal Nakshatras 
aad their bisection by cpluresu 


So far, wo have been going over the ground more or less 

traversed before by several scholars. But it may be asked 

if we have here reached the Ultima Thule of the Vedic 

antiquity. Does the oldest hymn, the first utterance of the 

Aryana mind, reach back thus far and no further ? Was it 

such a hymn that the Brahma- vadins of old and Panini several 

centuries before Christ believed and declared as c< seen"? 

. In what follows, I propose to bring together such evidence 

from the Vedic works as would enable us to deal with these 

questions. I have already drawn attention to the fact that 

the Ohitra and the Phalguni full-moon are mentioned as the 

mouths or the beginnings of the year in the passage from 

the Taittiriya Sanhita last quoted and discussed. In the 

next chapter I shall endeavour to show how these state- 

ments are to be interpreted, how far they are corroborated 

by other evidence and what conclusions we may deduce 




PhllgnnJ f all-moon, the new year's night S^yana's explanation unsatis- 
factory Phllguna could not "be a Vasanta month Two-fold character 
of the seasons, lunar and solar, superfluous Discussion of a passage 
in Shushruta Bh^skara Bhatta's explanation Winter solstice on the 
full-moon in PhSlguna The position of other cardinal points Vernal 
equinox in Mrigashiras AgraMyant Katiye Lexicographers' explana- 
tion of the word Grammatically objectionable Its real meaning 
according to Pan ini Erroneous rank of M<Lrgashtrsha amongst months 
according to the Bhagavad G!fc and Amara Margashlrsha could not 
have been the first month of the solistitial or the equinoctial year- 
It leads to the libration of the equinoxes Possible reason of the 
librafcion theory Mrigashiras -AgraMyana or the first Nakshatra in 
the year Mula, its primary meaning Evidence of the summer 
solstice occuring in Bh&drapada Origin of the annual feasts to the 
manes amongst Hindus and Parsis Comparison of the primitive 
Hindu and Parsi calendar Summary of results. 

THE passage from the Taittiriya Sanhita quoted in the last 
chapter states that the Chitra and Phalguni f ull-moons were 
the beginnings of the year, which then commenced with 
the winter solstice in the month of Mlgha. The words used 
in the original are Chitrd-piirna-'mdsa and Phalgunt-purna, 
-mdsa and these must be understood to denote, not the 
Ohaitra and the PMlguna months, whether sidereal and 
synodical, as Prof. Weber seems to have, in one case, sup- 
posed., but the full-moon days in each of these months. 
This is evident from the fact that these have been recom- 
mended as alternative times for the commencement of the 
satra in opposition to the Ek&shtak& day. In the case of 
the Phalguni-pirna-rndsa we are further told that Yishuv&n 
counted from that time falls during the rainy season, and it 




is impossible to suppose that Vishuvan can be counted from 
a month. The whole context, therefore, shows that it is a 
discussion as to the particular day best suited to commence 
the yearly sacrifice, and that Chitrd-purna~mdsa and Phal- 
guni-punia-mdsa must mean the days when the moon is full 
near the asterisms of Chitra and Phalguui. In the Tanclya 
Brahmaria* Phalguni-yurna-rndsa is rendered by Phdlguni 
and Jairnini has paraphrased Chitr&-pdrna-mdsa by CJudtoi 
and Phalguni and Chaitri, according to Panini (iv. 2. 3)., are 
the names of days. These interpretations have beem apeept- 
ed by all the Mimansakas including SSyana, and we may 
do the same especially as there are several passages in 
thje Taittiriya Sanhita where .pdrnQ-mfea is usd-m. a, similar 

But why should the OhiM and the: Phalguni 
be called the beginnings of the year ? Siyana thinks that 
they were so described because they occumred during 
Vasanta or the first of the seasons 4 But the explanation 
.does not appear satisfactory. I have previously shown that. 
.according to all astronomical works Shishira, commence^ 
^ wj^teir^ solstice, and/ that tKe tibiae seasons of Shi- 
ma were pomprisedi the Uttarfcyana 
N,0^ in the days of the Taitti- 


- quoted in the last chapter. 
IL 2. 10. 1. we find m^^^ simikrly used. 
WT are mentioned together ,-. while m ii. &. 4r. 

are contrasted. 
commeirtary on Tattt. San. -nL 4. 8 f . speaking of 

' and 

. . . . 

' 8*7?a observes 3W^*tf?f^^ 


Me irf. 


r!ya Sanhita the winter solstice, as shown in the last 
chapter, fell in the month of Magha ; and Magha and 
Phalguna were therefore comprised in Shishira, and Chai- 
tra and Vaishakha in Vasanta. But in order that S&yanVs 
explanation might be correct Phalguna must fall in the 
Vasatita season which, as a matter of fact, it did not. In, his 
commentary on the Baudh&yana Sutras" 34 " and also in the 
Kalamadhavat S yana tries to get over this difficulty by 
proposing a double Vasanta lunar and, solar, the lunar ta 
include the months of Phalguna and Chaitra, and the solar 
those of Chaitra and Vaishakha, quoting amongst others, 
Rig. x., 85, 18, as an authority to show that the seasons 
were regulated by the mo.on. The authorities, however, are 
not explicit and therefore sufficient to maintain the two-fold 
character of the seasons ; nor do I see the necessity of the 
two-fold character. It is true that the months in the: 
calendar were all lunar, but the concurrence of the lunar 
and the solar year was always secured by inserting an inter-, 
calary month whenever necessary. Under such a system 
lunar seasons can have no permanent place. Now and then 
lunar months ceased, as they now do, to correspond with the 
seasons they represented, but this was at once set aright by 
the introduction of an intercalary month. If we, therefore, 

* The passage is quoted in India: what it can teach us ? p. 32$ 
Sayana there quotes Taitt. San. vii. 4. 8., and after noticing 
that the CMtra and the Phalguni full-moon are both said to begin 
the year, he observes : 3T*T3T W^^HI^R" TOrT: 3*?f 3T 

frr **$ \ & 

The theorjrof the two-fold seasons thus appears to have 
been started simply to reconcile the two statements about the Cbitrl 
and Phalguni full-moons. , 
, f See Cal. Ed., pp. 60, 61, . ... J .,^^-'t 


exclude the correction due to the precession of the equi~ 
noxes, which was too minute to be noticed till after hundreds 
of years, there was thus no reason why the lunar seasons 
should come to be regarded as a permanent institution. But 
even accepting Sayana's two-fold character of the seasons, 
it can be easily shewn that it does not support his conclu- 
sions. A lunar year is shorter than a solar year by 11 
days. If the solar Vasanta, therefore, commences on the 
3 st day of the lunar Chaitra month this year, it will com- 
mence on the 12th day of Chaitra (lunar) next year and 11 
days later still in the third year when by the introduction 
of an intercalary month the commencement of Vasanta will 
be again brought back to the 1st day of Chaitra. The two- 
fold character of the seasons may thus delay the beginning 
of Vasanta to Vaishakha (lunar), but the season cannot be 
accelerated and brought back to Phalguna. It is true that 
in the days of Sayana (14th century) Vasanta commenced, 
as it does now, in the month of Phalguna; but it was so 
because the winter solstice had receded by over full one 
month by that time. Sayana does not appear to have fully 
realised the reason of this change and combining the occur- 
rence of Vasanta in Phalguna in his time with the occurrence 
of the same season, in Chaifcra in the days of the Taitciriya 
Sanhita and other works he attempted to reconcile the 
difference on the theory of the two-fold character of the 
seasons. But we can now better understand the change as due 
ty* the precession of the equinoxes, and must, in consequence, 
reject Sayana's explanation as unsatisfactory. 

The only other authority I can find ' for supposing that 
Phalguna was a Vasanta month is the statement in Shu- 
shruta's medical work, that " Phalguna and Chaitra make 
Vasanfca." But on a closer examination of the 


wherein this sentence occurs, it will be found to bear on 
its face the marks of later insertion. There are two conse- 
cutive paragraphs in Shushruta, each enumerating and 
describing the seasons of the year. The first states that 
" There the twelve months, beginning with Magha, make 
six seasons, comprising two months each. They are Shi- 
shira, &c Of these Tapa and Tapasya make Shishira " 

and so on until all the six seasons in their usual order, the 
ay anas, the year and the lustrum are described; and at the 
end we have " this is called the wheel of time by some." * 
The second paragraph then begins with the words " But 
here" and continues to state "But here the six seasons are, 
Varsha, Sharad, Hemanta, Vasanta, Grishma and Pravrish," 
thus altogether dropping Shishira and dividing the rainy 
period into two seasons Varsha and Pravrish, The para- 
graph then proceeds to assign the months to the seasons as 
follows : "Bhadrapada and Ashvina is Varsha, Kartika and 
Margashirsha is Sharad, Pausha and Magha is Hemanta, 
and Phalgana and Chaitra is Vasanta;" and so on until 
all the months are assigned to their respective seasons. 
The second paragraph, however, makes no mention of the 
ayanas, the year, or the lustrum. It is therefore evident 

* See Shushruta, Sutrasthana Adhyaya 6. The two consecutive 
paragraphs here referred to are : 




that the writer of the second paragraph, whosoever he 
may be> wished to note that the seasons and their corres- 
ponding months mentioned in the first paragraph had ceased 
to represent the actual state of things in the writer's time 
and province, and not thinking it desirable or possible to 
expunge or correct the old paragraph, he added immediately 
after it a second paragraph describing the seasons as he 
saw them. The words *' but here " at its beginning, the 
assignment of four months to the rainy season, but under 
two different names of Pravrish and VarsM, to keep up the 
old number of the seasons* and the absence of any reference 
to the ay anas, the year and the lustrum described in the 
praTious paragraph all point to the conclusion that the 
Second paragraph is of later origin and inserted with a view 
only to note the changes in the occurrence of events de- 
scribed in the paragraph next preceding it. It might be 
contended that the second paragraph is that of Shushruta, 
who notices the old order of things ia the first. But I need 
not go into that question here. For in either case it is 
plain that the passage wherein Phalguna and Chaitra are 
argued to Vasanta is the production of a later writer, who- 
he may be, whether Shushruta or any one else, and 
present inquiry is concerned we cannot tak& 
'as an authority for holding that PhMguna was 
in the days of the Taittiriya Sanhita. I ma^ 
:, tiwtfe Vagbhata who professes to summarise 
and Chamka giTOs the order w& 
we find Jft in the first patagrapb 
aHudmg to the clianges nofeetJ in ibfr 
1 may, therefore, suppose thai 


the paragraph did not exist in Vagbhata's' time or that he 
did not regard it as genuine. 

There is thus no reliable authority, that I am aware of, for 
holding that Phalguna, in the days of the Taittinya Sanhitft, 
was a Vasanta month, and Sayana's explanation does not 
in consequence hold good at least in this case* The ex- 
planation is further inconsistent with the fact that in 
several Brahmanas and Sfttras the full-moon night in the 
month of Phalguna has been pronounced to be the first 
night of the year. The Shatapatha Brahmana (vi. 2. 2. 18) 
says "the Phalgunl full -moon is the first night of the year. 1 ' 
The Taittiriya (i. 1, 2, 8) and the S&nkhyayana (iv. 4 and 
v. 1 ) * Brahmanas contain similar passages, while the 
Gopatha Brahmana (i. 1 &) after stating that the Uttar 
and the Pttrva Ph&Iguni are respectively the beginning and 
the end of the year, adds e< just as the two ends of a thing 
meet so these two termini of the year meet together."f I 
have already quoted a passage from the Tandya Brahmana 
to the same effect. The Sutra^writers, though not so 
explicit, do however distinctly state that the annual sacri- 
fices "should be commenced either on the Chaitrl or the 
Ph&lguni full-moon night, "$ thus clearly indicating that 
these were regarded as the beginnings of the year. If 

* iqrerr ^ sfac^rcw wrr ^rf^nw^sg^ft <fHfarcfr Shat. Br. vi. 2. 


I Taitt. Br. i. l. 2. 8, 

l San. Br. iv* 4. 


shvalayana Shr. Su. i. 2. 14, 3 ; KAt. Shr . Su. v. 1. 1 ; Sin. Shr. 
Su.iii. 8. L, xiii. 18.3. 


these passages mean anything, we must hold that the Ph3l- 
gutn full-moon night was once considered to be actually the 
first night of the year, or to put it in a modern" form the new 
yew's night. We cannot assign this position to it by 
simply assuming, as Sayana has done, that the night occurred 
during the two months of Vasanta. Sayana it 
appears, was aware of this objection and so in commenting 
on the passage from the Taittiriya Sanhita, quoted in the 
last chapter, he attempts to explain the position of the 
Phalguni night by reference to the above mentioned passages 
in the Br&hmanas, while with respect to tlie Chaitri, he 
Cjuietly observes that " this too is the mouth of the year as 
il during the season of Vasauta." * But an explana- 

admittedly fails in one case must fail in the other, 
for the Chitrft and the Plialguni nights are described to- 
gather, in the mme passage and in the same words, as the 
beginnings of the year. 

It will be clear from the above, first, that the theory of 
tie lunar seasons, started by Sayana to account for the posi- 
tion assigned to the Phalguni night in the Vedic works, 
cannot ha?e a permanent place in the Vedic calendar ; 
even accepting the theory, the beginning of the 
fasanta might be put off to the month of (lunar) 
YmisMkia, font could not be brought back to any day in 
PhftlgBBa ; and thirdly, the express texts in the Brahmanas 
declaring the Phalguni full-moon to be the new-year's 
Bight are inconsistent with S&yana's explanation. We must 
therefore look for some other solution. 

Bui if Siyaiia's explanation cannot be accepted, a&.' least 

* See ifae original remark quoted mpra, The wc^i ** too y \ in 
this explanation implies that it holds good also in die-ease of the 
ii! foil-moon. 


with respect to the Phalguni night, how are we to 
interpret the several passages in the Sanhita and the 
Brahnaanas given above? We cannot suppose that the 
Phalguni full-moon commenced the year at the vernal equi- 
nox ; for then we shall have to place the vernal equinox 
in Uttara Bhadrapada, which to render possible in the pre- 
Krittika period we must go back to something like 20,000 
B. C. The only other alternative is to make the full-moon 
eommence the year, at the winter solstice, and from the fact 
that the Maghi, the Phalguni and the Chaitri full-moons 
are mentioned together in the same passage of the Taittiriya 
Sanhita, and for the same purpose, I conclude that this is 
the real meaning of the passage in the Taittiriya Sanhita 
and those in the Brahrnanas. It is the most natural and 
reasonable interpretation of the passage and I find that 
Bhaskara Bhatta, who is older than Sayana, fully adopts 
this view in his Bhashya on the Taittiriya Sanhita. "* I have 
however devoted so much space to the discussion of Saya- 

* A MS. of Bhaskara Bhatta's Bhashya on the Taittiriya 
Sanhita has been recently discovered at Mysore and through the 
kindness of Sir Sheshadri lyar, the Dewan of Mysore, I have been 
able to procure a copy of the Bhashya on the passage here discussed, 
Bhaskara Bhatta after commenting on the first part of the passage 
which states that the sacrifice should be commenced on the 
Ekashtaka day, makes the following observation as regards the 
alternative next proposed : qt ^f^^r ISTRTt <TK*Ff H% 1 

ff ^ *rrer 
ffir i BT 

I As regards the third alternative proposed 
in the text, viz. the Chitra fall-moon Bhaskara Bhatta observes 
further on: 3T3": t r$rTr!T ^Tf^Pflf^ 1 F^^T ! |^ I Tr < 6' fc^nf^^ \ 

\ Finally Bhaskara 

Bhatta follows Jaimini and Shabara in the interpretation of the last 
part of the passage and concludes by observing that the best time 
lor the sacrifice is 4 days previous to the full-moon in Magha, 




na's explanation as the "high authority of that scholar is 
likely to mislead us in the interpretation of the passage. 
The Bhashya of Bhaskara Bhatta fully shews that Sayana 
is not here following- any older tradition and the reasons 
given, by him for explaining the position assigned to the 
Phalgtmi full-moon in the Vedic works are mere conjectures 
and guesses of his own. I admit that even the guesses of 
a scholar like Sstyana deserve consideration. But when on 
a closer examination we find that they are not supported by 
any eld traditions and are besides objectionable on various 
other grounds, I think we are bound to reject them. A.B 
observed by Bhaskara Bhatta the passage in the Taittirfya 
S&tbit& must, therefore, be understood as referring to an 
older year beginning, and we must hold that the fall-moon 
in PMlguna did as a matter of fact once commence the year 
at the winter solstice. I know that this view has been 
regarded as improbable by some scholars, on the sole 
ground that it would, if substantiated, enhance the 
antiquity of the Vedic works by about 2000 years more 
Iban what these scholars are willing to assign to them ; 
aacl as the natural result of such prepossessions amongst 
fltept "She subject has till now remained uninvestigated. 
Bui I hope that they will patiently examine the evidence, 
dxrat ami corroborative, which I intend to put forth in 
support of the suggestion and then give their judgment 
upon it. There is. no a priori impossibility involved in the 
hypothesis that the old priests, after changing their, start- 
ing point to the Krilfcik^s and framing the calendar accfrl- 
* n gfyr continued t<5 recogniae f or sacrificial purport , tlte 
older positions of the Nakshatras, just as all B^ttytaft fe<W 
the Himalaya to He Oape Cpmorin at pre^etil 'p^tfetfti tlutfr 

Sacnfl/*,>Ara rm ^w rfv*-!,^ J$, 1 ir*^^i -*.* if __T_ +_,_ JT ^ . ,' 4 .I../ 1 . * A ^ * 


was in 'the Krittikas. I think the pres-ent Brahmans are 
worse off in this respect, inasmuch as. they have not even 
the liberty, which the passage in the Taittiriya Sanhita 
accorded, though hesitatingly, to the old priests, of choosing 
either the old or the new calendar. To use the words of 
Professor Max Miiller we must in s-uch cases, therefore, u keep 
our preconceived notions of what people call primitive 
humanity in abeyance for a time.," * and form our judgment 
of antiquity, as we do of other facts, solely upon evidence. 

We have seen in th>e last chapter that the evidence for 
placing the vernal equinox in the Krittikas consisted of (1) 
the lists of the Nakshatras all beginning with the Krittikas, 
(2) the winter solstice then falling in the month of Magha, 
{3) the Nakshatra at the summer solstice being presided 
over by the- pi'tris , and (4-,) the possibility of considering, as 
Bentley suggested, the portion of the Nakshatra at tfhe 
autumnal equinox as divided by the equinoctial colure. In 
short, if the year was supposed to have begun in the month 
Magha, the position of the four cardinal points of the ecliptic 
as referred to the Nakshatras, was consistent with, and 
so indirectly established the truth of, such a supposition. 
Let us see if we can produce similar 1 evidence for establish- 
ing the hypothesis (for it is no better at present) that the 
year in the old Vedic days began, as stated in the Brahma- 
nas 9 with the Phalguni full-moon, and that the winter solstice 
occurred on ihat day. 'On a rough calculation the vernal 
cquincix, must recede two divisional Nakshatras to make 
the seasons fall back by one month. If the winter solstice, 
therefore, occurred in the month of -Phalguna, one month 
in advance of Mfigha, in the old Vedic days, the vernal 
equinox must then have been in Mrigashiras or two Nak- 

* India : what it can teach us ? p. 112. 




shatras in Advance of the Krittikas. Taking tie data given 
in the Vedanga Jyofeha as hia basis, the late Krishna 
SMstri Godbole has thus calculated* the position of the 
four cardinal points of the ecliptic, when the winter solstice, 
as stated in the Brfihmanas, occurred on the full-moon day 
in the month of Phalguna : 

(1) The winter solstice in 3 20' of the divisional 
tiara Bhadrapada; 

(2) The vernal equinox in the beginning of Ardra; 

(3) The summer solstice in 10 of Uttara Phalguni ; and 

(4) The autumnal equinox in the middle of Mula ; 

or giving up the system of reckoning by the divisional por- 
tions of the Zodiac, we have, roughly speaking., the winter 
solstice quite near the asterism of Uttara Bhadrapada, the 
vernal equinox between the head and the right shoulder 
of Orion or about 3 east of Mrigashiras, -the summer sol- 
stice at a distance of within 2 east of Uttara Phalguni, 
and the autumnal equinox about 5 east of the asterism of 
Hula. If we suppose the vernal equinox to coincide with 
Mrigashiras, the three other cardinal points are brought 
nearer to the fixed asterisms, and this appears to bo tho 
more probable position of the equinoxes and the solstices 
in those days. But without entering into these details, it 
will "be evident from this that when the winter solstice fell on 
the Phalguni full-moon the vernal equinox must be very 
near the asterism of Mrigashiras or two Nakshatras in 
advance of the Krittikas. We have now to see what evidence 
there is in the Vedic works from which this old position of 
the four principal points in the ecliptic may be established. 
There appears to be no express passage in the Ve$ic 

* See his essay oiivthe Antiquity O f the ; Vedas,;p* 19. 




works, wliicli states that Mrigashiras, like the Krittikas 
was ever the mouth of the Nakshatras. But what is so 
lost may still be discovered, in the words of Prof. Mas Miiller, 
4 ' hidden in the secret drawers of language." Mrigashiras 
may not be specifically described as the first of the Naksha- 
tras ; but the word Agrahdyanl which Amarsinha (i. 3. 23), 
gives as a synonym for Mrigashiras, and which supplies, 
according to Panini, a derivative word for the month of 
Margashirsha tells the same tale. Agrah&yanl literally means 
" commencing the year ; " and the question is how did the 
Nakshatra come to be so called ? In explaining the forma- 
tion of this word all native lexicologists, begin by assum- 
ing that the full-moon in the month of Margashirsha was 
the first night of the year, hence called Agrahdyani, and as 
this full -moon occurred in the month of Margashirsha the 
montlWtself was called Agrahdyanika There is no gram- 
matical inconsistency so far. But when these lexicographers 
further tell us that the Nakshatra itself was called Agrahdyani, 
as Amarsinha has done, because the full-moon in the vicinity 
of that Nakshatra commenced the year in old days,* one feels 
that there is something wrong in this explanation. The ordi- 
nary course is to name the full-moon or any other day after the 
Nakshatra, as Chaitrl, Pausham, Paushi, &c. (Pan, iv. 2, 3), 
while in the present case the order is reversed and the 
Nakshatra, we are told, is named after the full -moan. It 
is true that the lexicographers were, to a certain extent* 
compelled to adopt such a course, as they could not other. 
wise explain why AgraMyani, a term usually denoting a 
full-moon night, should have been given as a synonym for 

* See Bhanu Dikshita's commentary on Amar. i. 3, 23. 
explains the word thus : ^ ^R^T^f'- 1 

1.1 irr ? r?T^ D fr ^V^TrBl* i 




the Nakshatra of Mrigashiras by Amarslnha. But what- 
ever their motive, we have now to see if their explanations, 
as well as the statement in Amara, are correct. Turning 
to Panini we find no authority for this converse process. 
The word Agrahdyam occurs in Panini iv. 2, 22, which lays 
down the rule that the derivative names of months are 
formed from Agrahdya?ii and Ashvaf-tha, by the addition 
ofthak, * as a necessary termination; and this gives us the 
words Agrahdyanika and Ashvatthika for the months of 
MargashSrsha and Ashvina. Now in the previous. 9&tr& 
(iv. 2.21) Panini states that the names of the months art 
derived from the names of the full-moon days that occur 
in those months. It appears, therefore, that he understood 
Agrahdyam to mean the full-moon and not the Nakshaira 
of Mrigashiras. The word Agrahdyanl occurs thrice in 
Panini (iv. 2. 22 ; 3. 50 ; and v. 4. 110) and in all places 
it denotes the full- moon day. It is not, however, clear 
whether Panini treated it as a word derived in the same 
manner as Ohaitn, &c. If we, however, rely on analogy 
there is every reason to hold that Agrahdyam, like Kdrtiki 
and Phdlguni, may have been derived from Agrahdyana, and 
that this may originally be the name of the Nakshatra of 
.Mrigashiras. This supposition derives support from the 
fact that if, like Aunarsinha, we take Agrahdyanl as 
synonymous with the Nakshatra of Mrigashiras and follow 
the native grammarians in deriving this name of the Nak- 
shatra from that of the full-moon, it is very difficult to 
account for the initial long vowel in Agralidyani* All 

*_The sutras of Panini referred to in this discussion are 

:. ( fr.'2. 3 ), ^rf^lr^nEfrflr snerprr? ( iv. 2. 21 ), 

( fr- 2. 22 ), fo$RTRrer ( iv. 1. 41 ) and 
. 38). 


iexicograsxhers derive the word form Agra and Sdyana 
combined in a Bahuvrihi compound and afterwards adding 
the feminine termination ; thus Agra ,-f- hdi/ana-\-L But 
the feminine termination cannot be added without a pre- 
vious suffix (an) which also gives the initial long vowe! 3 as 
I is not a general feminine suffix, but is only used in special 
cases. We cannot get this an by Panini iv., 2. 8, as 
Agrahdyaiia is not the name of a Nakshatra according to 
Amarsinha. Various suggestions have, therefore, been 
made to account for the initial long vowel. Bhattoji suggests 
that we should obtain the long vowel by including- 
Agrakdyana in the Prajnadi list ( Pan. v. 4. 33 ) ; but in. the 
(Janapatha^ the list is not said to be a c specimen list/"* 6 " nor 
is the word Agrah&yana specifically included in the list 
there given. Boehblingk and Both in their dictionary 
obtain the long vowel by Pan. v. 4. 36 ; but here 36 may 
probably be a misprint for 38. Taranatha in his Vachas- 
patya obtains the long vowel by Panini v. 2. 1Q2 9 Vartika 1; 
bat Jyotsuadi is not again expressly said to be a * specimen 
list/ Bhanu Dikshita, the son of Rhattoji, in his commentary 
on Amaraf adopts his father's view and refutes that of 
M mkmta. The latter obtains the initial long vowel from the 
very fact that the word itself is so pronounced by Panini 
in iv, 2. 22 ; but this gives us Agrakdyani as a ready made 
word at once, and Mukuta had to assign some reason why 
the word should have been again included in the G-auradi 
list in Pan. iv. 1. 41. Mukuta's explanation is that P&nim* 
thereby intends to show that the feminine termination in 
Agrakdyani is not dropped in compounds. But Bhann 

tj meaning that the list is not exhaustive. 
t See p. 62 of the Bombay Ed. of Bhanu Dikshita's com. on 


Dtkshita replies "by observing that tie Gauradi list was 
never intended for the purpose and that as regards the 
accent we can get it otherwise. Bhanu Dikshita's own 
explanation or that of his father Bhattoji also dispenses 
with the necessity of including the word in the G-aurftdi list 
as they obtain the feminine suffix i by Pan. iv. 1. 15; and 
so in -replying to Mukuta he observes at the end that the 
"inclusion of the word in the Gauradi list is questionable. 1 ' 
Thus if we suppose Ainarsinha to be correct *md accept 
either Bhattoji' s or Mukuta* s derivation of AgrahAyam we 
shall have to hold that the word in question was either - 
wrongly included or subsequently inserted ia the GraBradi 
list and that Panini, who knew the word, forgot to insert 
it in the Prajnadi or the Jyotsn&di list. Both the explana- 
tions are again open to the objection that in this instance 
the Nakshatra is named after the fall- moon as against the 
usual method given by Panini in iv. 2. 3. 

The whole of this difficulty, however, vanishes, if we give 

?up the notion, that the full-mocn night in the month of 

Margashirsha might have commenced the year at one time 

\ and that the name of the Nakshatra as given by Amara 

^tnust be derived from the name of the full-moon. There is 

no express authority in the Vedic works to support such a 

theory and a closer examination of P&nini's sutras points 

to the same conclusion. Months in the Hindu calendar 

receive their names from the full -moon nights occurring in 

them ; and the characteristics of a month are the same as 

these of the full-moon night after which it is named, if the 

full-moon night in Margashirsha was, therefore, eTer the 

Hy&w'-year's night then the month itself would have come to 

be properly called the -first: month of the year. III other 

words the month of Margashirsha wonld iteeM", in that case, 


be called AgraJidyana. Boehtlingk and Koth do interpret 
the word AgraMyana in this way on the authority of Shab- 
da-kalpa-druma and Taranatha has done the same probably 
on the same authority, for none quotes any passage where 
the word is so used. Now if Agrahdyana ever meant the 
mouth of Margashirsha, the word would also assume the 
form Agrahdyana on the ground given above by Bhattoji ; * 
and we shall have Agrahdyana as another name of the 
month of Margashirsha. The word occurs in the Gauradi 
list (Pan. iv. 1. 41), and therefore must be taken to have 
been known to Panini. What did he understand it to 
mean ? There is strong ground to hold that he could not 
have understood it to mean the month of M4rgashirsha. 
For if we suppose that in Panini's times there were two 
forms of the word in. this sense Agrahuyana and Agraha- 
yanika he would have rather mentioned Agrahdyani in iv. 
2. 23,-f along with Ghaitri, &c., which gives the double forms 
Chaitra and Ghaitrika and nob with AsJivattha in iv. 2. 22 f 
as he has now done. We may, therefore, infer that Agra-* 
Jidyanika was the only sanctioned form of the word to denote 
the month of Margashirsha in Panini's time. This means 
that P&nini did not know of the theory which makes the 
year commence with the Margashirshi full-moon night or 
the month of Margashirsha (AgraJidyana). Tf so, he could 

* Bhanu Dikshita, in his commentary on Amara i. 4. 14. gives 
Agrahdyana as a synonym for Margashirsha on the authority of 
Purushottama and obtains the initial long vowel by including the 
word in the Jyotsnadi list. 

t The sutras are : srriTfni 1 ^^^^^ (iv. 2. 22 ) npTF*T W- 
: (iv. 2. 23). As the sutras follow each 

other i is natural to suppose that Agrahdyan^ if it gave rise to 
two forms, would have been included in the second sutra. 


not have derived the word Agrahayani for the full-moon 
night directly by taking it to be a Bahuvnhi compound.* 
The only other alternative is to derive it as we derive 
Chaitri and other similar words, and I think this is what 
Panini meant. For if he had been aware of any such diffi- 
culty in the formation of Agrahdyani, a word thrice used 
by him, and especially in obtaining the initial long vowel 
as Bhattoji and others have felt by taking* it to be a 
Bahuvrihi compound, he would have naturally noticed it 
himself. I therefore conclude that Panini derived Agra- 
hdyant from Agrdh&yana, as the name of a Nakshatra. 
In this case we can derive Agrahayani in a simple and easy 
naanner. For by Panini iv. 2. 3, we get the initial long 
vowel, when derivative words are formed from the names 
of the Nakshatras to express time; we now want the femi- 
nine suffix i, and though this could have been obtained by 
Pan. iv. 1. 15, yet, for accentual purposes, it may be consi- 
dered as provided for by the inclusion of the word Agra- 
Myana-f in the Gauradi list in Pan. iv. 1. 41, We can thus 
derive the word in the ordinary way, and unless we have 
strong grounds to maintain that it was really the full-moon 
mght and not the Nakshatra, which commenced the year, 
we*shall not be justified in accepting unusual derivations 
and explanations of these words. It is true that the word 
Agrahdyana as denoting a Nakshatra is now lost and 
Amarasinha only gives Agrahdyani and not Agrahdyana as 

* For then the full-moon night, and hence the month, would 
itself be the commencement of the year. 

t Doubts have been raised as to the exact form of the ward 
mentioned in the Gauradi list, and Bhanu Dikshita goes so faft as 
to question whether the word was really included in, the *!ist by 
Panini. . 


a synonym for the Nakshatra of Miigashiras. But I shall 
presently show that Aniarasinha is not alone in misconceiv- 
ing the meaning of these old words. The theory that the 
Margashirshi full-moon was the first night of the year, has / 
been the source of many other errors in later literature ; 
but before examining these it was necessary to show how 
the theory has distorted the natural meaning and derivation 
of the very words on which it appears to have been based. 
As remarked above if there be any express or cogent autho- 
rity to support the theory we might connive at the etymo 4 - 
logical difficulties, bub if it be found that the theory is 
inconsistent with many other facts, or leads, as I shall pre- 
sently show, to absurd results, the etymological distortions 
would afford us an additional ground for rejecting it. 

We shall now examine in detail the theory tha-t the f ull- 
jnoon night in Margashirsha was once the first night of the 
year. So far as I am aware there is no express authority 
for such an hypothesis except the statement in the Bhagavad- 
Grita (x. 35) where Krishna tells Arjuna that he, Krishna, 
ia " Margashirsha of the months (and), Vasanta of the 
seasons. 1 ' Anandagiri in his gloss on Shankara's Bhashya 
upon the GIt, observes that Margashirsha is here specially 
selected because it is a month of plenty. But the reason 
does not appear to be either sufficient or satisfactory ; for 
the next sentence, and in fact the whole context, shows 
that Margashirsha was here intended to be the first of the 
months. The principal commentators on the Git& are too 
philosophical to notice this point, but in a commentary 
written by Siirya Pandit, an astronomer, entitled the 
Paramrtha-prap&, I find that he explains the statement 
on the ground that Margashirsha was otherwise called 
Agrahdyanika, and th latter word denotes that the. full- 




BKX>B night in this month was the first night of the year.* 
If we accept this explanation, and no other plausible one is 
iorftmmiBg, it appears that this statement in the Bhagavad- 
was based on an etymological misconception of the 
among of the term Agrah&yani'ka ; and later writers like 
lia and Yagbhata,t simply followed the Gita in 
the same position to the month of Margashirsha. 

We may, therefore, treat all these statements as coming from 

oae source and representing a certain period of the Sanskrit 

literature, when native scholars first misconceived the 

primary meaning of AgraMyanikct. I have already shown 

properly understood, the etymology of the word gives 

EteUazbora for such a misconception. Agrahdyanika is really 

a ietiwmtive word and cannot therefore mean that the 

denoted by it was the first in the year just as 

Jjeehtiia does not mean the eldest month. But it appears 

tfihftt the tradition about Mrigashiras (Agrahdyana) ever 

; feeing the first of the Nakshatras, was completely lost in 

I ifcoee days, and native scholars believed, on what they con- 

* Tte commentary is printed at Poona. The words in the 
are SffT 


. .,. ^ -,^ --.I If Anandagiri's 

fee correct then the Gita is not opposed to deriving 
war ._^J fbia Agrahdyana, the name of a Nakshatra, and 
tie of tk k>ve discussion would he unnecessary. 


f TigUi8|i^ k hfe larger work entitled Ashtangasangralm, 
otherwise IM|iiiIi& Vaghhata, enumerates the months as begfe- 
^^ ( ;^fc In i. 4 of the wrk the Uttarayaia m 

said to commence wit-k Magha, while Margashirsha is mentioned 
amongst tie months there enumerated, much after the same 
way as Amara has done in i. 4, 13: and 14. . ; , , 


sidered to be sound etymological grounds that tbe montt 
and not the Nakshatra was the commencement of the year. 
Once started and embodied in the Gita, the theory gained 
an easy and rapid currency amongst native scholars, all of 
whom naturally felt bound to shape their views accordingly. 

And not only literary scholars, but astronomers appear to 
have done the same. In old astronomical works the year 
commenced with the winter solstice atxl the first month of 
the year meant the first month of the Uttarayana whicfe. 
commenced, with this solstice. If then the Marg;ash1nshi 
full-moon was said to be the first night of the year, an as- 
tronomer would naturally understand such statement to 
mean that the winter solstice fell on the full-moon, day of 
Margashirsha. Now if we suppose that the Margashirshi full- 
moon was thus the night of the winter solstice,, it would 
mean that the full moon on that day happened to be near 
the asterism of Mrigashiras. With the sun at the winter . 
solstice, the moon, to be full, must be near the summer 
solstice; and therefore the summer solstice must have then 
coincided with the asterism of Mrigashiras. The verna-l 
equinox is 90 behind the summer solstice; and if Mriga- 
shiras coincided with the latter, the vernal equinox would 
then be 90 behind the asterism of Mrigashiras. This is 
the only logical and mathematical conclusion possible, if 
we accept the theory that the full-moon night in Marga- 
sMrsha was the first night of the year at the winter solstice* 
And what does it mean ? It means a clear mathematical 
absurdity to us, though older astronomers,, not realizing its 
full effect, invented an explanation to account for- iL The 
Surya Siddhanta (viii. 2. 9} gives 63 as the polar longitude 
of Mrigashiras, counting from Revati. Now if the Yeroal 
eqoiaox was 90 behind the aster ista. of Mrigashipas, it 


90 63=27 behind the asterism of Eevati ! * The Vedic 
works, on the other hand, mention the "Krittikas as the first 
of the Nakshatras, and the winter solstice is shewn to have 
then occurred in the month of Magha. This means that the 
vernal equinox must be placed at least 26 40', or nearly 27 
in front of Revati. Now imagine the position of the Indian 
astronomer, who could neither reject the statement in the 
Vedic works, nor the one in the Bhaevad Gita. Both were 
sacred and unquestionable tests, and it would be no wonder 
if, to his great relief, he got over the difficulty by proposing 
a libration of the equinoxes, 27 on either side of Revati! 
The hypothesis is now given up by modern astronomers as 
mathematically incorrect; but no reason has yet been as- 
signed why it found place in the Hindu astronomy. A 
theory may be erroneons, but even an erroneous theory 
cannot become prevalent without a good cause. It has been 
suggested by Bentley and approved by Prof. Whitney t, that 
the limits of the libration might have been determined by 
the fact that the earliest recorded Hindn year had been 
made to begin when the sun entered the asterism of Krit- 
toka or 26 40' in front of Revati. Bnt this alone is not 
enough to suggest the theory of libration. For, unless the 
Hindu astronomer had grounds to him conclusive and 

* This may imply that the Surya Siddhanta was in existence at 
the time when the libration theory was started. I think it was. 
But it has been suggested that the libration theory might have been 
subsequently inserted therein (see Whitney's Sur. Sid., p. 104). It 
is not, however, necessary to make any supposition regarding the 
existence of the Surya Siddhanta at this time, as almost all other 
Siddbantas give the same bkoga, viz., 63 for Mrigashiim See 
Colebrooke's Essays, Yol ii., p. 325 (table). 

t See Surya Siddhanta, p. 103. 


otherwise inexplicable for holding that the vernal equinox 
fell 27 on each side of Revati, lie would not have proposed 
the libration of the equinoxes. So far as I know no such 
grounds have been yet discovered by modern scholars, and if 
the explanation given above accounts for the theory in all 
its details, I see no-reason why it should not be accepted as 
a probable explanation. Perhaps, it may be asked, what 
grounds I have to suppose that the astronomers com- 
bined the two statements declaring that Magha and M&rga- 
shirsha were both, each in its turn, the first months of the 
year, and so obtained the theory of the libration of the 
equinoxes. This is, however, not the place to go fully into 
this discussion; for all that I am bound to prove, as far as 
the present inquiry is concerned, is that if we accept the 
theory that the M&rgashirsha full-moon was ever the new- 
year's night, it leads us to an absurd conclusion, and this 
is evident from the above whether it does or does not give 
the real explanation of the libration theory. I may, how- 
ever, remark that when we actually find Amarasinha first 
stating (i, 4. 13) that " seasons comprise two months each 
beginning with Magha, and three such seasons make an 
ayana?' and then in the very next verse enumerating the 
months commencing with. Margashirsha ; there is nothing 
extraordinary in the snpposition that some Hindu astrono- 
mers might have 'similarly attempted to reconcile what were 
then regarded as the two beginnings of the year, by placing 
the statements in juxtaposition and pushing them to their 
logical conclusions. On the contrary, I should have been 
surprised if the Hindu astronomers had not done so. 

Bufc, apart from the origin of the libration theory, I think 
it is clear that, if we accept that the Margashir&ha full-moon 
was ever a new-year's night, in the sense that the winlte* 


solstice occurred ai/ that time, wo are inevitably landed on 

an absurdity. By the ordinary process of rednrtio ad , 

absurdum, we are thus compelled to abandon the theory that ! 

the full-moon in Margashlrsha once began tho year at the 1 
winter solstice. Native scholars and astronomers, who did 
! not realize the -absurdity, accepted, tho -theory of the libra- 
1 tion of the equinoxes as the only ponsiblo way of reconciling 

i the two statements in their sacred books. Wo now know I 

| that the equinox cannot be placed 27 behind llovati, unless r f | 

I it be either in the beginning of tho present cycle of the I 
f precession of the equinoxes or about 000 years hereafter, f s 
i and we should have no difficulty in rejecting tho premises 

I I that give us such a conclusion. Perhaps it may bo urged ^ 
f that the full-moo a night in Mirgashlrsha might have boon }l 
I called the new-year's night in some other sonso,* Yes, it S| 

* The only other explanation, I know of, is that given by I 

Eentley in his Historical Survey of the Hindu Astronomy, pp. 5,-27. ! f 

Bentley divides the zodiac into 27 lunar mansions, beginning with ^ 

Shravishtha in the winter solstice, as in the Vednuga JyotiBha. I 

Then he divides it again into 12 tropical months beginning with if 

H&gha. The beginning of Magha and the divisional ShraviBhthft 
tims coincide at this time. Now the beginning of each month 
must &U back owing to the precession of the equinoxes ; and in 
thus* receding if the beginning of any month coincided with any 
fixed Itmar mansion, on the 6th lunar day, the month, says Bentley* 
was -made to commence the year! But what authority is there io 
native astronomical works for such an elaborate and artificial theory 
to determine the Commencement of the year ? Native astronomers 
are surely expected to know better the theory on which they 
commenced their year. Then, according to Bentley's calculations, 
A*lrvina was the first month in 1192 JBL 0. and .Kftrtika in 946 B. 
C. But there is no evidence whatsoever in the Sanskrit literature to 
, corroborate these resulit Again why should either of these month* 

I 1 




might be; but what evidence is there that any nativs 
scholars ever thought of it ? None that I know of. There 
are only two beginnings of the year known in ancient Hindu 
literature. I have shown that the winter solstice could not 
have occurred on. the full-moon in Margashirsha , and by 
the same method we can prove the improbability of the 
vernal equinox falling on that day. For if we suppose the 
Margashirsha full-moon to be the new-year's night, in the 
sense that the vernal equinox occurred on that date, we 
must make the asterisin of Abhijit coincide with the vernal 
equinox. This gives us about 20,000 years B. C. for the 
period when these positions could have been true. The 
author of the Bhagavata Parana appears to have had some 
such theory in his mind when he paraphrased (xi. 16- 11) 
the above quoted verse in the Gtta by " I am Margashlrsha 
of the months, Abhijit of the Nakshatras," and the late 
Krishna Shastri Godbole took this statement for a record 
of a real tradition ! This illustrates the danger of relying 
on traditions in later books, without tracing them to their 
source in the oldest works we possess. 

We must therefore rise above these etymological spe- 
culations of the native scholars of what Prof- Max Muller 

not have been called Agrahayavdlsa ? Bentley supposes that, this 
method was in force till 538 A. D. ; if so ? why should Pausha not 
become Agrah&yanilsa instead of Margashirsha, inj 451 B. C. ? 
Bentley J s unsupported speculation Doust, therefore, be rejected as 
imaginary* It gives no reason why Margashirsha, the third of 
the several months which, according to his theory, would successive: 
ly begin the year from 1192 B. C. to 538 A. D., should alone have 
been called AgraMy arnica* and none whatever why the Nakshatra 
should be called Agrahdyaw, contrary to the usual rule, according 
to which the word should denote the full-moon day. 


once called the Renaissance period of the Sanskrit li- 
terature. It is these speculations that have given us the 
libration theory and interrupted the tradition of Agrahdyana 
coming down to us iritact. It is difficult to say how these 
etymological speculations originated. Perhaps the word 
Agrahdyanika was in course of time corrupted by non- 
user into Agrahdyana on the analogy of Chaitra and Cliai- 
trika, and such corruption gave rise to these speculations, 
or it might be that the year locally commenced with 
Mirgashirsha in certain provinces, and attempts were made 
to find an authority for such custom in the etymological 
meaning of the word Agrahdyanika. It appears to me more 
probable, however, that the old tradition about the Nak~ 
shatra gradually got connected with the month which was 
named after it as in the case of Kartika, whose first rank 
amongst months is suggested by Prof. Whitney ff as due to 
the ancient position of the Krittikas as the first among the 
lunar mansions.' 1 * This is very likely if, as shown below, 
the word Agrahdyani was ever used to denote both the 
Nakshatra and t-he full-moon. But whatever the origin j 
the . speculation was there safe under the authority and 
prestige of the Bbagavad GHtft, and Amarasinha, who appears 
to have been not wholly free from the influence of such the- 
ories, naturally put down Agrahdyani instead of Agrahdyana, 
as the name of the Mrigashiras, especially as the latter 
word, Agrahdyana, was not expressly mentioned by Pmnu 
Later lexicographers, who considered Amara and especially 
the QM to be above error, attempted to reconcile Anmm's 
statement with the system of Panini by unusual derivations, 
and astronomers appear to have vied with them in ! mathe^- 
matically reconciling the real and the imaginary "begin- 

* See.his Surya Siddhanta, p. 271 (xiv. 16 n^ 


nings of the year ! We must, therefore, set aside all these 
theories and go back to the purer times of Panini, to deter- 
mine what was the real name of the Nakshatra. I have 
already shown that Panini knew the word Agrahdyana and 
also that he could not have understood it to mean the month 
of Margashirsha. It is, therefore, evident that he used it 
as a derivative from Agrahdyana in the sense of time as 
given in Panini iv. 2. 3. If so, he considered Agrahdyana to 
be a name of the Nakshatra of Mrigashiras. Amarsmha's 
Agrahdyani is, therefore, either an error or a feminine 
adjective for the tdrd of Mrigashiras meaning exactly the 
same thing as Agrahdyana: thus Agrahdyana = Agrahdyana 
(Pan. v. 4. 38), Agrahdyana-\-i (Pan. iv. 1. 15)= Agrahdyani* 
In support of this derivation, may be cited the fact that 
Mrigashiras was once considered to be a feminine word, 
Mukuta and Bhanu Dikshita t both quote, Bopalifca who 
gives the neuter and the feminine forms of Mrigashiras. 
Raman&tha in his Trikanda Vivaka, gives a quotation from 
Babhasa and another from a Smriti to the same effect.^ If, 
the word Mrigashiras was thus ever used in the feminine 
gender, the feminine adjective Agrahdyani might have 
been used as a synonym for the same, not because it 
was the name of the full-moon, but because the asterism was 
spoken of in the feminine gender. This may account for 

* This is open to the objection that we hare to include Agra- 
hdyana in the Prajnadi list. 

f Amara i.3. 23. Bhanu Dikshita's commentary is printed in 
Bombay and Mukuta's and Kshirasvamin's are published in 
Atmndoram Borooah's unfortunately incomplete edition of Amara 's 

See extracts from Ramanatba's com. in Anundoratn Borooah's 
, p. 112. 



the fact why Amarasinha lays particular stress on this point. 
For says he " Mrigashlrsham (is) Mrigashiras ; Agrahdytmi 
(is used) to denote the very some* ; 1} thus implying that ai 
feminine word is nsed to denote what he supposed might be 
regarded only in the neuter gender. This is, indeed, a plau- 
sible explanation. It not only absolves Amarasinha from, 
the charge of having given a wrong, or at least a distorted, 
word, but makes him warn his readers not to misunderstand 
the word Agrahdyanl for the full-moon night -a mistake 
into which almost all his commentators have, however, unfor- 
tunately fallen. It may further explain why instead of the 
Nakshatra, the full-moon day (both of which were on this 
theory denoted by the same word Agrahdyani) came to be 
regarded as the firsl night of the year and so gave rise to 
later speculations. But the fact that Amarasinha mentions 
MUrgasMrsha first amongst the months shows that he was 
not altogether free from the influence of the speculative 
theory ; and the explanation above stated must therefore be 
accepted with caution. 

But whatever explanations we may adopt to defend 
Aniara, I think it will be plain from the above that, so far as 
our purpose is concerned, we must reject the explanation 
dtibe commentators of Amara, who derive the name of the 
HftkBhatra, as given by Amara, from Agrahdyani, the name 
of the foil-moon. After this we may either suppose Agra- 
hdyana or Agra/idyanl or Agrahdyani to be the name of the 
IJakshalra, for in every case the difference consists only in 
the form and gender and not in the derivation, or. the mean- 
ing of the word. Thus understood Agrahdyani or Agra- 
Jidyana both give us the same meaning, viz., that tke yvar 
was in Hi front of the Nakshatra of Mrigashiras ; or in other 

* Thus : 


words commenced .with it. If what I have said 'above is 
enough to prove this, I do not care to insist on a particular 
-form, whether masculine, feminine, or neuter, oiAgrahdyana 
which as an adjective is the basis of all such forms. With 
this reservation, I may, I think, in what follows use the 
word Agrahdyana to denote the Nakshatra of Mrigashiras 
and as evidencing the circumstance that it was so called 
because it was the first Nakshatra in the year. 

Corresponding to the winter solstice in Phalguna, we 
thus have the asterism of Mrigashiras or Agrahdyana to 
commence the year from the vernal equinox, much after 
the same manner as the Krittikas were said to be the mouth 
of the Nakshatras when the winter solstice fell in the month 
of M%ha. The express statement in the Brahmanas that 
the PMlguni full-moon commenced the year from, as I have 
previously shown, the winter solstice, is thus borne out by 
the tradition which we find treasured up in Agrahayani. 
Now if the vernal equinox was near the asterism .of Mriga- 
shiras the afutuinnal equinox would be in Mula. It has been 
ingeniously suggested by Bentley that this name signifying 
** root or origin" may have been given to the Nakshatra 
because it was once the first amongst the asterisms and he 
Bias actually given a list of the Nakshatras beginning with 
Mula; bui he does not appear to have used it except to 
show that when one of the twenty-eight Nakshatras was 
dropped the divisional Jyeshtha and Mula both began from 
the same fixed point in the heavens, a position which gives 
him the vernal equinox in the beginning of the Zodiacal 
portion of the Krittikas, I have already shown that we 
cannot suppose that the old Vedic priests made observations 
of imaginary lines in the heavens, and Bentley's explanation 5 
wbi0h entirely, .depends, oil tie mathematical divisions of 


tlie Zodiac is not therefore satisfactory. Nor can I accept 
Prof. Whitney's suggestion that Mula " may perhaps have 
been so named from its being considerably the lowest or 
farthest to the southward of the whole series of asterisma 
and hence capable of being looked upon as the root of all 
the asterisms. 1 "* I should rather suggest that Mula was so 
called because its acronycal rising marked the commence- 
ment of the year at the time when the vernal equinox was 
near Mrigashiras and the winter solstice fell on the Phal- 
guni full-moon. Agrakdyana setting with the sun in the 
west and Mftla rising in the east then marked the beginning 
of the year, and this position of Mula is likely to be especially I 

noised as the heliacal rising and setting of a star/ and so of ;* 

Agrakdyana, is difficult to be accurately watched. The Jj! 

etymological meaning of Mula may thus be said to supply ] 

a sort of corroborative evidence for placing the vernal 

eqainox in Mrigashiras though, in absence of other "strong $ > 

grounds, it is of no better valu than a similar conjecture of 5 

Bentley about the name Vishakha, noticed in the last t *\ 

chapter, ' ,/ 

I have already mentioned before that the year was di- " 

vided into two ay anus, the northern and the southern, and f 

that though originally the northern ay ana indicated the 
passage of the sun to the north of the equator yet it afterwards 
came to indicate the passage of the sun from the winter 
to the summer solstice. I have also stated that after this 
change was made all the attributes of the older ay anas 
must have been gradually transferred to the new ones, 
though the old division was concurrently kept up ; and 
that the new ideas were formed solely with reference to the 
solstitial division of the year. Thus the Pitriyftna luring 

* See his Surya Siddhanta, p. 194, 


which time tlie stm in older times went down the equator 
must have come to be regarded, for some purposes at least, 
as commencing from the summer solstice. With the winter 
solstice occurring on the PhalgunJ fail-moon day, we shall 
have the summer solstice on the Bhadrapadi full-moon, 
so that the dark half of Bhadrapada was the first fortnight 
in the Pitriyana, understood as commencing on the summer 
solstice. It was thus pre-eminently the fortnight of the 
pitris or the manes ; and to this day, every Hindu celebrates 
the feast to the manes in this fortnight. As far as I know 
no reason has yet been advanced why the dark half of 
Bhadrapada should be called the fortnight of the pitris 
(pitri-paJcslia) and why special feasts to the manes should 
be ordained at this particular period of the year. With the 
winter solstice in the asterism of TJttara Bhadrapada, that 
is when it occurred on the PhsUgunl full-moon, the matter 
is simply and satisfactorily explained. For then the Dak- 
shinayana or summer solstice commenced on the dark half 
of Bhadrapada and this fortnight therefore naturally be- 
came the first fortnight in the ayana of the manes.* 

And not only the Hindus but the Parsis celebrate their 
feast to the manes at the same time. The coincidence is 

* This explanation implies that the feast to the manes became 
permanently fixed at this time ; and there is nothing improbable in 
it. For as the Parsis hold similar feasts on corresponding days 
we must suppose that these feasts became fixed long before the 
Parsis and the Indians separated. When the vernal equinox receded 
to the Krittikas the feasts still continued to be celebrated in the dark 
half of Bhadrapada. But though the priests could not alter the days 
of these feasts, yet in assigning duties to the Nakshatras they re- 
^tgniaed the change by making pitris preside over Magha at tha 
summer solstice. 


important inasmuch as we are here dealing with periods 
of antiquity when the Indian, the Iranian, and the Hellenic 
Aryas must have lived together, and if our theory is correct 
it is sure to be corroborated by the customs, practices, and 
traditions of the other two sections of the Aryan race. I 
shall in the next two chapters show that there is ample 
independent evidence of this kind confirmatory of the theory 
that Mrigashiras commenced the equinoctial year in those 
early days. At present I shall only refer to the conclusions 
of Dr. Geiger as to the nature of what he calls the primitive 
or the oldest Avesta calendar. He takes madhyaryo which 
literally means not * mid- winter/ but 'mid-year' as his 
basis and concludes that in the primitive Avesta calendar M 
the year commenced with the summer solstice.* This is / 
just what we should expect. The Indian Aryans com- V 
menced their year from the winter solstice or the beginning [ 
of the Uttar&yana and the Iranians, who in such matters 
always took a diametrically opposite view, naturally com- 
mented it with the summer solstice the beginning of the 
Dakshin&yana, thus bringing the Sruma (or the winter* 
solstice) in the middle of the year. But the coincidence 
does not stop here ; and in the light of the old Indian | 
qalendar we are in a position to explain some difficult points 
in the primitive Avesta calendar. The Hindu pitri-paksha 
or the fortnight of the manes commenced with the summer 
solstice, while the Iranians celebrated their feasts to the 
manes just at the same time. The first month in their 
calendar was called Fravashinam or the month of the manes, 
according to the primitive calendar determined ,'lbjf 

* See Dr. Geiger's Civilization of the Eastern 
Ancient Tinies, translated by Darab Dastur Pheshtpteatt 
Vol. I., p. 153. , 

if\ > 



Dr. Geiger, this first month, when the feasts to tne mane a 
were celebrated,* began with the summer solstice. Again, 
the fourth month of the Avesta calendar was Tishtryehe or 
the month of Tis try a, which has been identified with the star 
Sirius. Counting with Bhadrapada in the summer solstice, 
the fourth month in the Hindu calendar would be Margashir- 
sha or the month of Mrigashiras, which Nakshatra is 
quite near Sirius. We can now also easily explain why 
Dathusho should have been dedicated to the Creator. 
Beginning with Fravashinam in the summer solstice, 
Dathusho begins exactly at the vernal equinox, and as mark- 
ing the revival of nature it was properly dedicated to the 
Creator. Both again was partially correct when he ima- 
gined that Dathusho must have once commenced the year 
inasmuch as it was dedicated to the Creator Ahuramazda. 
For from, the old Hindu calendar we see that the vernal 
equinox was also a beginning of the year. In the primitive 
Avesta calendar we can thus discover the traces of the year, 
beginning with the vernal equinox and also from the sum- 
mer solstice (in opposition to the Hindu winter solstice) 
in BMdrapada, the month of the manes. These coinci- 
dences, especially about the month of the manes, cannot be 
said to be merely accidental. The worshippers of Ahura- 

- * The last five days of the old year and the first five days of the 
new year are called 'fFravardigan" days. "During these ten days 
the frohars (fravashi or fravarti ) the spiritual representatives of 
the deceased are believed to come to the houses" of men on the 
earth. See Dr. Hang's Essays on the Parsis, p. 225 note. At pre- 
sent the Hindu feasts extend over the whole of the fortnight. We, 
however, find an alternative period recorded in the Nirnaya Sindhu, 
which states that the feasts may extend over a fortnight, ten days 
0r five days ! 


changed the commencement of tho year from .th 
winter to the summer solstice, but as observed by "Kotli cf a 
sacred and solemn feast conld not be removed from its 
place in the year/** and this affords therefore a comparatively 
reliable ground to identify the Avesta and the Vedic year. 
We find nothing in the Avesta to explain why the first 
month of the year should have been devoted to the manes ; 
but, as observed by Dr. Geiger in respect of the legend of 
Tama, the knowledge of it might in course of time have 
been lost to the worshippers of Ahuramazda. We can, how- 
ever, now easily explain it from the statement in the Vedic 
works that PMlgun! full-moon was once tie new year's 
night at the winter solstice. I know that such analogies 
taken singly are of no great practical value, but when from 
& consideration of the Yedic literature, we arrive at results, 
which we then find so similar to those arrived at independ- 
ently by Zend scholars, we may certainly be led to believet 
that they are not merely accidental. 

To sum up : Interpreting the passage in the Taittiriya 

* See Dr. Geiger's Civ. An. Iran., Vol. I., p, 145. The annual, 
totfee manes amongst the Parsis came after the Gahanbartf 
and It is interesting to note that the pitri-pdkska is defined in the 
Strya SildMata, xiv. 3-6, as the period of 16 days after the feur 
or festivals at intervals of 86 days each begin-. 
vitfc libra. The author of the Surya Siddhanta is here 
mfcnilj describing some old festivals and as Edshis were in use 
Is iw days k fees the duration of these festivals according- to the 
fafeadbr ^B In force. The mention of Libra does not therefore 
pawt us fern regarding Skadashiti-mukkas as old festivals. Bmt 
wtietiiw fiUdEadttMwUM were in any way connected wltla ilia 
k is a0i etsy to determine in the present tafce of 00* 


Sanhita,, which, states that the "Phalguni-purna-mdsa is 
the mouth of the year/' in the natural way suggested by the 
context and similar other passages in no less than five 
Brahmanas, to mean that the winter solstice occurred on the 
Phalguni full-moon in those days, we find that Mrigashiras 
has been designated by a name, which, if properly under- 
stood, denotes that it was the first of the cycle of the 
Nakshatras, thus showing that the vernal equinox was ; 
once near it; that Miila can now be better understood as the 
star that rose at the beginning of the first night of the 
equinoctial year ; and finally the fortnight after the summer 
solstice was devoted to the feast of the manes, as the ayana 
of the pifris commenced at that point ; and that this is fully 
corroborated by the Parsi month of the manes falling in 
their primitive calendar at the same time. It was on 
evidence like this that the old position of the Krittikas was 
determined, and I do not see why a similar conclusion about 
Mrigashiras should not be allowed It is true that no | 
express statement has been cited to show that Mrigashiras \ 
commenced the cycle of the Eakshatras in those days and that * 
some scholars may not consider the evidence of Agrahdyani 
sufficient for the purpose. In the following chapters I 
hope to show that there are a number of other circumstances 
and even express texts which leave little room for 
cautious fears like these. 



Mr%Mhiiw-Its oldest form and position Identification of Rohin! and 
Kadra, etc. Plutarch on the non-Egyptian origin of Orion, Canis and 
Una Methods of interpreting mythological legends Storm anec| 
dawn theories Their insufficiency Knowledge of the heavens amongst , ,, 
the ancient Aryas Heaven and Hell, Devayna and Pitriyana Joined! ' 
by equinoxes, the gates of Heaven Dogs at these gates Kerberoar } 
Yu&a'fl dogs The Chinvat bridge and the dogs that guard it '* >< 
Ebeir identification with Canis Major and Canis Minor, when th,6h < i 
vmwl equinox was in Orien Celestial river and Charon's boat; ! 

of the Sigveda and the A vesta dogs SaramA and Shunfi-* 
; (star) commencing the year Heliacal and acronycal rising** , ^ 
spring and antnmn Vishnu, and Rudra Kerberos an$, <; 
The legend of Namnchi alias Vritra His decapitation by ? 
mt tbe gates of heaven, where Orthros is stationed Represented ',< 
fcjr tic * antelope's head " in the heavens, Vritra being = Mriga 1 - ' ^ r 
betwn Indra and Namnchi Watery foam Its identification, " < 
ibe Milky Way Legends of Rndra How he killed Praj&patah " \ 
ia or Sftcriic at the beginning of the year Shulagava sacrifice ' f ; ^ 
tri-9tri t the three-star belt of Orion The Hindu Trinity, Dsittii- * * v*l 
to^^r His representation in the sky. '^ ;^i ' 

,1 / (4 <! 

pnrt of the fceaveus, which contains the Nakshatras^lf / 
to consider, is the most attractive and interest-" , ^ 
JM6 celestial sphere. Even a casual observer 

is to be attracted by its splendid 
1 fP An of the sun in this 'portion of the 

m Hf of it year must have rendered it 

to Aryan observers. It conimB3' 

*** ^*^ f tibe jBhcsT m^nitnde iEid 1 
* *^*" f ** s^ondj with the sta-asto &p 
Wy ttcB^ Here thTO n^ ^ Atf 

- - - 

4 & 





priests and the numerous legends that exist in almost all 
the sections of the Aryan race about this portion of the 
heavens fully show that they did not fail to make use of 
this brilliant opportunity* I intend to examine some of 
these legends in this chapter with a view to see what 
corroborative evidence we may get therefrom. If we can 
more naturally and easily explain the legends that relate to 
this part of the heavens on the present theory;, than has 
hitherto been done, we may fairly conclude that we have 
rightly interpreted the passages from the Brahmanas; if 
not, we shall have either to revise our assumption or to 
give it up entirely. But before we do so we must 3 as far as 
possible, try to identify the asterisms and determine their 
forms as described in the ancient works. 

We shall first take up Mrigashiras or Agrahayani accord- 
ing to Amarasinha. The very name of the Nakshatra^ which 
means " an antelope's head/' * suggests the figure of the 
asterism. Bat the constellation consists of so many stars 
that it is very difficult to say which of them might have 
suggested the name. I may here remark that the doctrine 
of "Yogatar&s" or the junction stars cannot be supposed 
to have been developed in the early days we are here 

* I may here, once for all, remark that though I have translated 
the word Mrigaskiras by the " antelope's head," I do not mean 
to imply that Mriga necessarily meant " an antelope " in the Yedic 
literature. It has been suggested that Mriga may mean "a bullock" 
or some other animal like it. It may, hut we have nothing to do 
with it, inasnmch as the word Mriga itself is still used in the 
Sanskrit literature to denote the constellation. My translation of 
Mngashiras must therefore be considered provisional, remem- 
bering that though it may change yet the argument in this chapter 
will still remain unaltered. 


i t 

speaking of. I do not mean to say that single stars iriay ; 

not have been or were not specifically named. But where- !; 

ever a constellation is spoken of, it is more probable that I 

the whole group was intended, as in the case of the Seven | 

Bears or the Krittikas ; and hence the determination of 
the junction . stars, as given in later astronomical works, 
cannot help us beyond indicating where we are to look lor 
the constellation described in the old works. For instance, 
if we take Mrigashiras we are told that one of the three 
small stars in the head of Orion is the junction star. This 
means that we must look for Mrigashiras in the constellation 
of Orion. But how can these three stars give us the figure 
of an antelope's head ? The three stars are so close that 
between themselves they give us no figure at all. It is, 
however, suggested that the two stars in the shoulders and 
two in the knees of Orion give us the four feet of the 
antelope, whose head may then be said to correspond with 
the three stars in the Orion's head. In short, it is the 
antelope's head in the same way as it is the head of Orion* 
Bat besides being open to the objection that this gives us 
the head and not the form of an antelope's head, the 
explanation presupposes that the whole of the antelope is 
in the heavens ; and if Ardra be correctly identified with 
the- star in the right shoulder of Orion we shall have also to 
include this star in the four feet of the antelope. The old r 
Vedic works, however, seem to lay down that it was the 
head of the antelope and not the antelope itself, that was 
transplanted to the heavens. Keferring to the legend of |I 
Eudra piercing Prajfi-pati, S&yana in his commentary on the 
/ij 2/1/2. 8)* observes that he, tlie 

Sayaaa's commentary on Stoat. Br. iL 2. 1; 2. 

y.] THE ANTELOPE* s: HE AD. n 

terrible form created by the gods, (f cut off Prajapati' s head 
by the arrow," and "the arrow and the head both jumped up 
to the heavens and are there stationed." The Aitareya 
Brahmana (iii. 33) gives the same story and there too 
Pashuman or Bhutavan is said ta have pierced Prajapati 
with an arrow. But it does not distinctly say whether it 
was the head or the body that was pierced by him though 
in the Shatapatha Brahmana Mrigashiras is described as the 
head of Prajapati. The tradition of piercing the head does 
not, however, occur in this form in the Rigveda, though in 
Rig. x. 61. 5-7 this story of Prajapati is alluded to. But 
in many places where Indra is mentioned as killing Vritra 
we are told that he cut off the head of his enemy (i. 52. 10 ; 
iv. 18. 9; viii. 6. 7) and in Rig. v. 34. 2 and viii. 93. 14, 
Indra's enemy is described as appearing in the form of an 
antelope. This shows that the Rigveda indirectly speaks 
of an antelope's head having been cut off by Indra, and it 
may justify us in holding that Rudra did the same. The 
tradition is preserved even in the Greek mythology which 
tells us that Apollo, indignant at lier sister's affections for 
Orion, made her hit, with an arrow, a mark in the distant 
sea, whfeh turned out to be the Orion's head.* In the 
heaven^ we must therefore look for the " cut off" head of 
Mriga'with the arrow pierced in it. There are other 
circumstances which point to the conclusion. Sanskrit 
writers have described a small group of stars in Mrigashiras 
called Invakas. Amarasinha tells us that they are " on the 
top of Mrigashiras.f Now if Mrigashiras itself be under- 

* See Smith's Dictionary of Classical Mythology. Ov. Fast, 
v. 537. 

t Thus : ^rRfH? ^ I I'^K^IT^I *NH 3 f ^fi J^ ^ **! * 
JnT^TT RWM ^ I .Amara. i. 3. 23. ^nsf^ft^T 
according to Bhanu Dikshita. 




stood to denote the three small stars in the head of Orion, 
Invak&s become identical with them and the distinction 
given in Amara must be put aside as meaningless. I am, 
therefore, of opinion thab the asterism of Mrigashiras was 
once really believed to possess the/orm of an antelope's 
head with an arrow sticking to it. The mention of the 
arrow in these traditions at once enables us to determine the 
form, for the arrow can be readily and easily identified 
with the three stars in the belt of Orion. The head with 
the arrow at the top must therefore be made up by taking 
along with the belt the two stars in the knees and one in 
4he left shoulder o Orion somewhat as below : 

res us the arrow pierced into the heard and the 
j fi^rs in the belt are at' the top of tjie antelope's head 
. Amara assigns to Invak&s. I may further 
^ancient observers could not and would not 
a:e three small stars in the Orion's head to 

V f ' 

i. wlien there were so many stars of ibe 
in the s&tne portion of the 
the later ^sfeonqmers miay 
bte three stars 




those in the belt and not in the head of Orion. I do not mean 
to imply that the asterism may not have been conceived and 
figured otherwise. As a matter of fact we know that it 
was figured as a hunter or a deer/ and there are good grounds 
to hold that these are ancient ideas.* All that I, therefore* 
mean is that of the various figures we may make out of the 
stars in the constellation of Orion, one should be of an 
antelope's head with the arrow sticking to it to represent the 
cut off head of Mriga, and not as the present configuration 
supposes both the body and the head of Mriga together and 

I have in what has gone above presumed that the asterism 
of Mrigashiras must be looked for in the constellation of 
Orion, and that the legends of Eudra and Prajapati refer 
to this constellation. Some scholars, however, doubted the 
correctness of this assumption ; and so far as ' absolute cer- 
tainty is concerned their doubts maybe justifiable. For, 
Vedic hymns were not committed to paper till a long time 
after they were 1 sung, and there is of course no possibility of 
finding therein a photograph of the portion of the heavens 
referred to in the various hymns. All that we can, therefore, 
do is to weigh the probabilities of the proposed identifica- 
tions ; and if this course be adopted I do not think any 
reasonable doubts could be entertained about the identi- 
fication of Mrigashiras with the constellation of Orion. To 

* The constellation appears to have been variously conceived : 
(1) the antelope's head ; (2) the whole antelope ; (3) Prajapati 
either in tte fbrm , of an antelope or as a person with a belt or 
YajnopavUa (see the next chap.). Of these three forms I consider the 
* antelope's head ;*' tb be the ^gdeSt. It will be seen that the three 
forms are closely connected, and that they are the developments of 
the same idea. 


quote the words of Prof. Whitney '* there is the whole story, 
illustrated in the sky : the innocent and the lovely Rohini 
(Aldebaran) ; the infamous Prajapati (Orion) in full career 
after her, but laid sprawling by the three-jointed arrow 
(the belt of Orion) , which shot from the hand of the Bear 
avenger (Sirius) is even now to be seen sticking in. his body. 
With this tale coming down to us from the first period of 
Nakshatras in India who could have the least doubt of its 
persistent identity from the earliest times to the latest/'* 
I subscribe to every word of what is here so forcibly ex- 
pressed. Of course, we may expect some variations of 
details as the story got degenerated into Puranic legends ; 
but it is impossible to mistake the general identity. I shall 
therefore not unnecessarily dwell upon it here. 

We have seen how Mrigashiras may have been primitively 
conceived. After this it is not difficult to identify the 
other stars. The Rohini is no other than Aldebaran. Rudra 
is the presiding deity of Ardra, and we may therefore sup- 
pose Rudra to be represented by the star in the right 
shoulder of Orion ( a ). But the Aitareya Brahmana (iii. 33) 
identifies Rudra with Sirius or what} is now called the 
Mriga-vyadha. The Milky Way does not appear to have 
received a specific name in these old days, and the threes 
sections of the Aryan race the Parsis, the Greeks, and the 
Indians have no common word to denote the same. Yet 
it is impossible to suppose that this broad stream of stars 
could have been unnoticed, and I shall show further on that 
it was not. Greek Astronomy places two dogs in this parti 
of the heavens Canis Major and Canis Minor one on eacH 
side of the Milky Way, and it has been doubted whether 

* See Prof. Whltney'a-Bssay on Hindu and Chinese systems of 
asterisms, p. 53. * - - 


the claims of these dogs to primitive antiquity are well 
founded. In wliat follows, I hope to show that they are. 
In the meanwhile, I may here refer to the testimony of 
Plutarch to prove that some at least, of the actually existing 
figures of constellations in the heavens are Greek transforma- 
tions of others which had been placed there before by the 
Egyptians ; for this writer, who in his treatise De I side et 
Osiride makes the priests of Egypt say that the souls of 
gods shine in the heavens and are stars, adds that fc the 
constellation of Isis is called, by the Greeks, Canis ; that 
of Ebrns, Orion,, and that of Typhon, Ursa/'* This state- 
ment is very important, inasmuch as it shows that the 
names of at least three constellations, Orion, Canis and Ursa, 
are not of Egyptian or Chaldean origin. Of these Ursa 
Major (Greek ArMos) has been already identified with sapta 
rik&has or simply the rikshas of the Vedas and the Hapto-' 
iringa of the Parsis, thus partly confirming the above- 
mentioned statement of Plutarch; and it can be shewn, 
that his observation is equally good in respect of the other 
two constellations, or that Canis, Orion and Ursa are all of 
Aryan origin. At present I use Plutarch's statement only 
so far as to justify us in presuming the three constellations' 
to be of Aryan origin, or, to put it negatively, not borrowed 
by the Greeks from the Egyptians. t 

* De Iside et Osiride. I take the quotation from Narrien's 
Origin and Progress of Astronomy, p. 44. Narrien further 
observes that this assertion of Plutarch seems to be confirmed by 
the discovery of a sculptured planisphere on the ceiling of the 
Temple of Denderah where "in the place of Canis Major is traced a 
cow, the animal consecrated to Isis" and "instead of Orion is the 
figure of man which is supposed to be intended for the son of Osiris." 

t I have deemed it necessary to make these remarks because 
Mr. Gladstone in his Time and Place of Homer, p. 214, observes 
that Orion is either non-Hellenic or pre-Hellenic." Plutarch's 




Having thus shown that we are at liberty to assume that 
the Greek legends about Orion and Canis are not of foreign 
origin., let us see what coincidences we can discover 
between the legends of the three sections of the Aryan 
race about this part of the heavens. I am not going to 
trace every legend to its primitive source and explain it on 
the dawn or the storm theory. Nor do I believe that it is 
possible to do so ; for there are many other objects in 
nature besides the dawn and the storm, that are likely to 
impress the mind of a primitive man;* and a legend, thought 
it might have originated with the sun or the dawn, is sur& 
to grow and develop under the influence of these objects,: 
For instance* we can understand the story of Vritra by 
supposing that he represented the power that locked up 
the waters in the clouds, but when we are told that this 
Vritra sometimes assumed the form of a Mriga, here is a 
distinct addition which cannot be satisfactorily accounted 
for on the original theory. Those that have watched and 
examined how legends grow can easily understand what 
I mean. The idea that everything must be reduced to ' 'dawn 
and nothing but the dawn" is the result of supposing that 
in the days of the Rigveda men were not acquainted with 
anything else. The supposition is partly true, but as I shall 
presently show there are many passages in the Rigveda which 
presuppose the knowledge of stars and constellations. Thus 
at the time we are speaking of several ideas had already been 

testimony shews that the constellation is not of Chaldean or 
Egyptian origin. The conception must therefore be pre-Hellem*c, 
or, in other words, Indo-Germanic, and I think I have given ample 
evidence in this chapter and the next to prove that the i(}ea of 
Orion was fully developed before the Greeks, the Parsis and ^le 
Hindus separated. 

* See Herbert's Speaker's Sociology, Vol. L, 


formed and recognised and even familiarly known. For 
example, the idea of Devaydna and Pitriydna appears to 
have been well settled at this time, so much so that I. though 
the year was afterwards made to commence with the winter 
solstice, the equinoctial division of the heavens, with all the 
notions which had already become associated with it, con- 
tinued to exist, though somewhat restricted in its scope, 
side by side with the new system. Whether this idea itself 
is or is not further resolvable into simpler ideas is a ' 
different question altogether. Perhaps it may be shown 
to have grown out of the idea of day and night or light and 
darkness. There are several passages in the Rigveda (L 
123. 7; 164. 47 e ) which speak of a black and a white day, 
and it is very likely that these were the original names of 
Devaydna and Pitriydna; for when new ideas are introduced 
it is usual to express them in old words with such qualify- 
ing adjectives as would distinguish the new idea from the 
old one. A " black day" might thus mean the Dakshina- 
yana or the Pitriyana, as night appeared to increase at the 
expense of day during the period. When the southern 
course of the sun thus came to be likened to a dark day or 
l night (Ri'g. vi. 9. 1) it was naturally regarded as a night of 
the Devas to distinguish it from the ordinary night ; and as 
no sacrifices were performed during the ordinary night, so 
no offerings could be made to the Devas during their night 
(vi. 58. 1 ). Of course, it must have been a long time before 
men could develop conceptions like .these,, There was, 
indeed, a time when they could hardly account for the fact 
how the sun found his way from the west back to the east. 
In the Rigveda x. 72.7, the sun is said to rise from out of 
the ocean and a similar , idea is found in Homer who 
describes not only the sun, bat even the stars, as '*. bathed 


in the waters of the ocean."* In the Bigveda x. 108. 1, 
Saraina is said to have crossed really a " long way." The 
Aitareya Brahmana iii. 44, which states that the sun never 
sets in reality, makes a distinct advance upon these notions. 
But it is difficult to say whether astronomical ideas were 
developed to such an extent in the days when the year first 
commenced from the winter solstice. I do not, however, wish 
to enter here into these details. As -previously observed I 
assume that," at the time we are speaking, the Vedic Aryas had 
already passed through these stages, and that the ideas of 
Devaydna and Pitriydna were familiarly known and estab- 
lished ; and assuming these as established, I intend to examine 
how legends were built upon them. I have, however, briefly 
alluded to the probable origin of these ideas inasmuch as 
it helps us to better appreciate the description of the Deva- 
yana and the Pitriyana. Ordinarily the Pitriyana is 
described (Rig. ix. 113. 3 ) as the region " where Vaivas- 
vata is the king, which is the undermost (lit. obstructed t ) 
part of the heavens, and where there are eternal waters.-" 
The Vaivasvata Yama here spoken of does not, however, 
appear to have as yet been invested with the terrible charao- 
Rs ter we find given to him in the later literature. Corres- 
ponding to Tama in the south we have Indra in the north, 
each supreme in his own sphere, and dividing the whole 
world into two parts, one bright and known, and the other 
watery and mysterious, or, in the language of seasons, first 

* Lewis, Hist. Survey of the Astr. of the Ancients, p. 6. Iliad, 
v. 6, vii. 422. 

t sprracNsf f^T: in the original. I think S^fnFT means " when 
the view of the heavens is obstructed;" "the portion of the 
heavens which is turned away," Of. Ait Br, iv. 14, where 
of the year is spoken of. - 


comprising Vasanta, G-rishma and Varsha and the, second 
Sharada, Hemanta and Shishira* 

Now when the vernal equinox was in Orion or Mriga- 
shiras 'it was the beginning of the Devayana, and as the 
constellation is remarkable for its brilliancy and attractive- 
ness the ancient Aryans may have been naturally influenced 
not merely to connect their old traditions with it, but also 
to develope them on the same lines. Thus the Devayana 
and the Pitriyana, as representing the two hemispheres 
must be joined, and the vernal and the autumnal equinoxes 
became the natural points of union between the regions of 
gods and Yama. The equinoxes were, in fact, the gates of 
heaven, and as such it was natural to suppose that they 
were watched by dogs. In the Rigveda i. 48. 15 the dawn 
is spoken of as illuminating the "gates of heaven/ 5 and in 
i. 13. 6 andii. 3. 5 the gates-deities are invoked to keep the 
gates open. We have a similar invocation in the Vajasa- 
neyi Sanhita 21. 49. This shows that the idea of the "gates 
of heaven" was not unknown in Vedic times, and the ar- 
rangement of the gates on the sacrificial ground, which is pre- 
pared on the model of the annual passage of the sun, shows 
that these gates divided the whole hemisphere into two 
parts. Macrobius records a tradition that " the ancients 
designated the signs of Cancer and Capricorn as the gates 
of the sun, at which having arrived, the luminary seemed to 
retrace his path in the zone which he never leaves."* Now 
Macrobius could not but speak in the language of the 
twelve zodiacal portions, and if we therefore divest his 
statement of the form in which it is naturally expressed it 

> ,* Macrob. Comment, in Somn. Scrip. Lib, I. cap 15. I take the 
quotation from Narrien's Origin and Progress of Astronomy, p. 51. 


means that the equinoxes, which the ancients supposed to 
be once in the zodiacal signs named above, were then called 
gates of heaven. 

The Iranians, however, have preserved the legend more 
fully. With them the equinox is not merely a gate, but a 
bridge connecting heaven and hell the Devaloka and the 
Yamaloka, or the Devayana and the Pitriyana and "dogs 
that keep the Chinvat Bridge )} help the departing soul to 
cross it. Darrnesteter, in his introduction to the 'Vendidad, 
published in the Sacred Books of the Bast Series, observes* 
that "this reminds one at once of the three-headed Kerberos, 
watching at the doors of hell and still more of the four-eyed 
dogs of Yama, who guard the ways to the realm of death" 
(Rig. x. 14.10). The ideas are, indeed, strikingly similar 
and point ont to a common source. Kerberos has even been 
identified with Sanskrit Shdbala or Sharvara, meaning 
variegated or a dog of Yama, But, as far as I know, no 
satisfactory explanation has yet been given of these legends 
nor any attempt made to explain them on a rational basis.t 
If we, however, suppose that the vernal equinox was once in 
Orion, the constellations of Canis Major and Canis Minor 
the two dogs would then be on the boundary line of heaven 
and Yama's region, and the whole of the above story may 
be seen illustrated in the sky like that of Prajapati and 
Budra previously referred to.} According to Bundahis, 

* Sacred Books of the East, Vol. IY., Zend-Avesta, Part I., 
Introduction v., 4. 

f See Kaegi's Rigveda, by Arrowsmith, p. 160, note 27% 
where the writer quotes Aufrecht to the same effect. 

J Weber and Zimtner appear to have suggested that the 
conception of Yama/s dogs might have been formed f roin some 
constellations* BloomteW rejects this suggestion and tries to show 


xu\ 7, the Chin vat Bridge extends from tlie Height o- 
ChaHd-i-Daitak in the middle of the world to the summit 
of Arezur at the gate of hell ; while Dr. Geiger observes 
that "it was believed to have been built over a wide expanse 
of water which separates the paradise from this world, 5 '* 
In the later Indian literature we are told that the souls of 
the deceased have to cross a streamt before they reach the 
region of Yama, while the story of Charon shews that even 
the Greeks entertained a similar belief. What could this 
river be ? With the vernal equinox ia Orion, one can 
easily identify it with the Milky Way, which could then 
have been appropriately described as separating the regions 
of gods and Yama, the Devayana and the Pitriyana, or the 
Northern and the Southern hemisphere. In the later 
Hindu works it is actually called the Celestial River (svar- 

that the dogs represent the sun and the moon. His explanation 
does not, however, show how and why the dogs came to be located 
at the gates of heaven and why they should he entrusted amongst 
all the sections of the Aryan race with the duty of watching the 
souls of the dead. Bloomfield quotes Kath. S. xxxvii. 14 (where 
day and night are called the Dogs of Yama) and Shat. Br. xi. 1. 
5. 1. (where the moon is said to be a divine dog) to prove that the 
dogs must be understood to mean the sun and the moon. But I 
think that the Brahmana here gives simply a conjectural explana- 
tion, and, as in the case of Namuchi's legend, we cannot accept it, 
inasmuch as it does net give any reason why the dogs were station- 
ed at the doors of Yama's region. There are many other incidents 
in the story which are not explained on Bloomfield's theory. I see, 
therefore, no reason for modifying my views which were put down 
in writing before I could get Bloomfjeld's paper in the last number 
of the Journal of the American Oriental Society. 

" * Dr. Geiger's Civil, of East Iran., Vol I., p. 100. 

t Called Vaitarani The Garuda Purana, Pretak, vi. 25-31, 
states that a cow should be given to a Brahman to enable the 
deceased to pay the ferrymen on this river. 9 . .: 


nadi), while the Greeks have placed near it the constellation 
of Argos (ship) and two dogs Canis Major and Canis 

Minor one on each side to guard both the entrances of the 

Ghinvat Bridge across it. The Rigveda also mentions two 
dogs of Yama kept to "watch the way," while the Greeks 
place a three-headed dog at the gates of hell. In Kig. x. 
63, 10 we are farther told that the land of the blessed is to 
be reached by " the celestial ship with a good rudder." * 
The words in the original are daioim navam. Comparing 
these with the expression divyasya sJiunak in the Atharva 
Veda vi 80. 3, and seeing that a celestial (divya) representa- 
tion of Rudra is described in later worksf it seems to me 
that we must interpret the epithet to mean " celestial " and 
not simply " divine." Thus the Vedic works appear to place 
a celestial dog and a celestial ship at the entrance of the 
other world, .and these can be easily identified with the 
Greek constellations of Argo Navis and Canis, if we suppose 
the Milky Way to be the boundary of Heaven in these days. 
I do not mean to say that these conceptions had their origin 
in the appearance of the heavens. On the contrary, a com- 
parison with the n on- Aryan legends shows it to be more 
likely that the heavenly bodies received their names from 
the pre-existing beliefs, about the other world, amongst the 
people* .Herbert Spencer tells us that amongst the non- 
Aryan savage races the journey to the next world is believed 
to lie over land, down a river or across the sea, and that in 
consequence the practice of burying their dead in boats 
prevails amongst some of them.J The North Americans, 

*SeeKaegi's Rigveda, translated by Arrowsmith, p. 159, note 273. 
f See the passage from the Mahimna Stotra quoted infra. 
J See Herbert Spencer's Principles of Sociology, Vol. I., chap. 
XF., 1st Ed.. ....... 


we are farther told, say that the Milky Way is fe the e Path 
of Spirits j* * the Road of the Souls/ where they travel to the 
laud beyond fche grave, and where their camp fires may be 
seen blazing as brighter stars."* This coincidence between 
the Aryan and the non-Aryan legends makes it highly 
probable that the figures of the constellations were conceived 
by the Aryans according to notions of the next world 
prevailing amongst them at that time. It may be noticed, 
however, that the non-Aryan races do not connect the idea 
of time, e. </., of the year and the seasons, with these beliefs, 
while it is the chief characteristic of the Aryan legends. 
We are, for instance, told that the dog commenced the year 
(Rig. i. 161. 13) and that the Devayana comprised the 
three seasons of Vasanta, Grishma and Varsha (Shat. Bi\ ii. 
L 3. l).f It is this feature of the Aryan legends that is 
most important for the purpose of our enquiry, while the 
coincidence, above pointed out, confirms, in a remarkable 
way, the genesis of the Aryan legends here proposed. The 
chief elements in the traditions of the three Aryan nations 
may thus be satisfactorily explained. 

It may, however, be contended that the two dogs of Yama 
spoken of in the Rigveda may not be the same as the 
Avesta dogs at the Bridge. A closer examination of the 
several passages in the Rigveda will, however, dispel such 
doubts. In the Vendidad xiii. 9, the dogs are called peshu- 
pdna, or those that guard the way to the region of death. 
The Avesta dog is chat/iru-chashmen (Ved. viii. 16), while 
the Vedic dogs are described as chatur-akshau (Rig. x. 
34. 11), both of which expressions mean u four-eyed/' The 
dogs in the Avesta and the Rigveda, however, differ in colour. 

* Principles of Sociology,- -Vol. I.-, chap. x\iv., p. 399, 1st Ed. 
t For German legends, indicating time, see the next chapter. 




In Ved. viii. 16 the dogs are said to be zairitem or spaetem 
yellow or white with yellow ears ; while 


the dogs of Yaina are said to be shabalau, spotted or 
variegated. But the difference is neither very material, 
nor such as cannot be accounted for. In the Rigveda we 
can trace the yellow colour of the Avesta dogs. The 
antelope of the sun in Rig. x. 86. 3 is said to be harita or 
.yellow, the zairelem of the Avesta, and if we 'suppose this 
antelope to be no other than that represented by Orion, as 
the sun commenced the year at that point, we need not be 
.surprised if the dogs in the Avesta are described as yellow, 
especially when in the Atharva Veda viii. 1.9. we find the 
two messenger dogs of Yama named as Shydma and 
.Shabala, thus noting probably a difference in colour. The 
Atharva Veda iv. 20. 7 mentions a four-eyed bitch, while in 
the Shatapatha Brahmana xiii. 1. 3. 7 the adjective is applied 
to a dog ; and the same animal is evidently intended in both 
places. In the Parsi scriptures the dogs at the Chinvat 
.Bridge are sometimes spoken of in singular (Ved. viii. 16) 
and sometimes, as in Rig. x. 14. 11. in dual (Ved. xiii. 9). 
This shows that we might disregard gender and number in 
,the description of these dogs; and we are thus led to 
suppose that Sarama in the Rigveda is again to be identified 
with the dogs that watch the gates of heaven. Whether 
Sarama* in primitive days was or was not connected with 
the dawn, I do not undertake to say. But there is an, 
incident in her story which confirms the identification I 
have proposed. The Panis tried to coax Sarama by offering 
.her milk which she drank. On her return she denied 
having seen the cows of ludra, who thereupon kicked her 

* See Max Mailer's Lectures on the Science of Language, Vol 
II.,. p. 511. 


and she vomited the milk. Now the mention of milk at 
once suggests the idea that it must be the milk in the 
galaxy on each side of which the two dogs are stationed . 
In Rig. iv. 57. 5 Shunasirau are invoked in order that they 
may pour down upon the earth the C( mil k, " which they 
"make in. heaven*" Prof, Max Miiller records a suggestion 
that Shunasirau, here spoken of, may be a very old name for 
the Dog-star, and with its derivative Sairya would give us 
the etymon of Seirios I * In Rig, vii. 55. 2 the Vastoshpati, 
"the guardian of the house/' in the form of a dog, is 
invoked and described as bright and red Sdrameya on 
whose jaws spears seem to glitter: a description which 
answers so well with the appearance of Sirius, that with 
what has been said above we may at once identify the 
S&rameya with the Dog-star. I may here refer to the 
Shatapatha Brahrnana ii. 1. 2. 9, where speaking of Mriga- 
shiras, the Prajapati's body pierced by Rudra is described 
as his vdstu. May not Vastoshpati be regarded as guardian, 
of this? If so, it may be a further proof that Vastoshpati 
represents the star Sirius, which, as it were, guards the head 
of Prajapati in the" form of Orion or the antelope's head. 
But, apart from this suggestion, I would finally quote Rig. 
i. 161* 13, where it is expressly stated that "the dog 
awakened" the Ribhus, the genii of the seasons, at the "end 
of the year I" Sayana proposes to interpret shvdnam in 
the original by "wind," but it is evidently an error. In. 
the Shatapatha Brahmana xiii. 5. 1. 8, vrika and sTivd are 
mentioned together, and the former is known to be a name 
for a wild dog. If so, Sayana's explanation of Rig. i. 105. 1 1 
appears to be more probable than that of Yaska. It is in 

* Max Mailer's Lectures on the Science of Language, Vol. II,, 
p. 526.. . . ... , . v 



fact a description of the dog (star) appearing in the east 
after crossing "the eternal wafers V* of Yamaloka, and tlien 
being immediately lost in the rays of the sun, which rising 
after it, had to push the wild dog out of his way. The 
mention of the "eternal waters" of the Yamaloka indicates 
that the heliacal rising of the Dog-star, here referred to* 
occurred at the end of the Pitriydna or at the vernal equinox, 
thus further confirming the statement that the dog com- 
menced the year. There are -other passages of similar 
import, but as I wish to avoid, for the present, any disputed. 
passages, I do not mention them here. If the time, I am 
contending to establish for the hymns of the Rigveda, comes 
to b accepted, it is sure to furnish an unerring clue to tlie 
interpretation of many other passages and legends in that 
sacred book, but the work must be left to be done hereafter. 
Patting all these passages together, we find that in the 
Rlgveda, dogs are described as dark and brown, bright and 
red, possessing four eyes, guarding the house and the way 
to Yama's region, vomiting and making milk, and above 
all beginning the new year.* All these facts clearly show 
thai the Vedic dogs are the same as the Hellenic or 
the Iranian, and we can easily and satisfactorily account 

* Prdl Bloomfield's theory leaves many of these facts un- 
explained. If the dogs represent the sun and the moon, how can 
Hie SUB tell the Ribhus that the dog awakened them at the end 
of the year 7 I cannot also understand how the sun and the moon 
em be described as variegated in colour, or as engaged in makiBg 
milk. Agala how can the sun or the moon be said to be four- eyed, 
and why should they perpetually remain at the boundary of 
&eaen and faeM ? In Rig. x. 86. 4, a dog is said to be let 
kwse at the ear of the Mriga, and this as well as the dog in R%. i. 
161. 13, most be supposed to be different from Kama's togs, if ^e 
accept Bioomfield's Yiew. 


for all these legends by supposing that the vernal equinox 
was near the Dog-star in those days,, thus making the dog 
rise with the sun in the beginning of the year at the gates 
of the Devayana. "We can now also understand how the dogs 
could have been described as four-eyed. For, if they are 
correctly identified with Canis near the Milky Way, then 
the four stars in the body of Ganis might naturally be said 
to be his eyes ;* for once the number of eyes is increased 
from two to four, we need not expect to find them .all on the 
head, but, like the thousand eyes of.Indra in the later 
mythology, they may be regarded as spread over the whole 
body. M. Darmesteter rightly observest that "the Parsis 
being at a loss to find four-eyed dogs interpreted the name 
as meaning a dog with two spots above the eyes; but it is 
clear that the two-spotted dog's services^ are only accepted 

* In Rig. x. 127. 1, the stars are said to be the eyes of night. 
The Greeks entertained a similar idea. Their Argos was surnamed 
Panoptes, *' the all-seeing/' having a hundred eyes on the body. 
See Max Muller's Science of Language, Vol. II., p. 416. 

| Sacred Books of the East Series, Vol. IV.; Zend A vesta, Part 
I.; Intr. v. 4 

J These services are required at the funeral ceremony. It may 
be here noted that the hymn in the Rigveda which describes 
Yama's dogs ( Rig. x. 14 ) is still recited at the time of burning 
the dead body of a Hindu, Every Brahman has also to give, every 
day, two small offerings of cooked rice to the two dogs of Yama, 
Shyama and Shabala, at the time of the Paiskvadeva sacrifice. Several 
deities receive their oblations at this sacrifice. The offerings are 
placed on the ground in the form of a circle, beginning with the 
eastern point. The offering to Shyama is placed outside the circle 
at the south-west and that to Shabala at the north-west point. In 
other words, Shyama and Shabala are placed cm each side of the 


for want o a four-eyed one, or of a white one with yellow 
ears. 35 Evidently the Parsi priests failed to realise that it 
was the divine or heavenly, and not an earthly dog that 
was here described, as driving the death-fiend. The 
Atharva Veda vi. 80* 3 shows that the Indian priests of the 
time well understood it to mean a dog who is " born of 
waters, whose house is in the sky, and who sheds his lustre 
all around. " 

There is another set of traditions which we can similarly 
explain on the supposition with which we have started, viz., 
that the vernal equinox was then in Orion. The heliacal 
rising of the constellation at the beginning of the year 
marked the revival of nature at the commencement of spring, 
and the asterism may thus be said to represent all these 
milder influences which iu later mythology were fully 
embodied in the conception o Vishnu. But the case was 
completely reversed if we take the acronycal rising of the 
same. It was at the autumnal equinox that the Dog-star 
rose ab the beginning of night, and though, strictly speaking, 
it marked the end of Varsha, yet the portion of the heaven 
wherein the constellation is situated could have been easily 
regarded as the battle-ground of Indra' and Vritra who fought 
ib. those days, and also as the stage on which the terrible 
Eudra made his appearance. In short, the constellation 
naturally became the harbinger of the mild and the terrible 
aspects of nature. It is in this latter sense that the Dog- 
star might be considered a rain-star, and Sarama, like the 
Greek Hermes with which it is identified, might be said to 
have been sent to search for the cows of Indra taken away 
by the Pa^isof the nether world. The Greek legends mention 

' - , , ^ ^..^,^,,.4-1, ,',:;- UJ^ -. 

western point, in the s&rae way as the dogs appear in the heavens 
on each side of the Milky Way. 


two watch dogs Kerberos and Orthros ; and of these 
Kerberos has been etymologically identified with Sharvara 
and Orfchros with Vritra* But no explanation has been 
given of how this Vritra came to be stationed at the gates of 
hell. Prof. Max Miiller suggests that Orthros is the dark 
spirit that is to be fought by the sun in the morning. But 
then, this does not explain why it was called Vritra, and 
how it came to be killed by Herakles". The legend - of 
Namuchi, as given in the Rigveda and interpreted on the 
supposition that the year began with, the Dog-star, does, 
however, solve the difficulty. I have already alluded to the 
fact that in the Rigveda Vritra is often said to appear in the 
form, of a Mriga (Rig. i. 80. 7 ; v. 82. 3 ; v. Si. 2 ; viii. 93 . 
14). In Rig. vii. 19. 5 Vritra and Namuchi are both said 
to be killed by Indra, and though this cannot be taken as a 
direct authority for holding that Vritra and Narnuchi are 
the different forms of the same enemy, yet from the 
description of the two I do not think there can be any 
doubt as to their being identical. In fact, Shushna, Pipru, 
Kuvaya and Namuchif are only so many different names 
of the enemy of Indra. Now Indra is represented as 
cutting off the . head of Vritra (Rig- i. 52. 10 ), and also of 
Namuchi (Rig. v. 30. 7; vi, 20. 6). Combining these state- 
ments we get that Indra cut off the head of Vritra or 
Namuchi, in the form of a Mriga; and this at once suggests 
the question whether that head is not the same as that of 
Prajapati cut off by Rudra and which gave the name of 

* Max Mfiller, Clifford Lectures, 1891, p. 248. Biographies of 
"Words, p. 197. 

t See Prof. Bloomfield's contributions to the Interpretation of 
the Veda in the Journal of the American Oriental Society, Vol. 
XV. ? p. 146. 


Mriga-shirsha, or cf the antelope's head" to the constellation. 
In Rig.i. 53. 7, we are simply told that Namuchi was killed 
by Indra in the distant (pardvati) region, which seems to 
mean the region of Yama. But as it does not satisfactorily 
determine the place where Namuchi was killed, I refer to 
Rig. x. 73. 7, where Indra by killing Namuohi is said to have 
cleared np "the paths leading (ydna in the original) to the 
(region of) Devas;"* which plainly shows that Namuchi was 
killed at the gates of the Devayana. In the Vajasaneyi 
Sanhita 10. 14 a sacrificial rite is described which gives the 
same place and time of Namuchi's death. The priest there 
throws away a piece of metal hidden under a tiger "hide, 
exclaiming, " the head of Namuchi is thrown away/' after 
he has taken his Yajamdna through all directions (East, 
South, West, North and upwards) and also through all the 
seasons (Vasanta, Grishma, Sharad, Varsha and Hemanta- 
and-Shishira), This means, if it can mean anything, that 
Namuchi alias Vrifcra was killed, in the language of seasons, 
affcer Shishira, or in other words, at the gate of the Devayana 
as described in the above quoted passage from the Rigveda, 
for the end of Shishira is the eod of the Pitriyana. Here 
tfaea we have an explanation of how Orthros came to be 
afe && gate of hell, or in a distant region under the setting 
smn. But the association of Orthros with Kerberos throws 
farther light on the subject. If Vritra's head is the same 
as Mrigarskfrsha, as explained in the beginning of this 
chapter* .then .the three stars in the belt of Orion, which 
form the top of Mrigashiras, might have easily suggested 
the idea of a three-headed monster. In Rig. x. 99. .6 

* The original verse is as follows : 


Indra is said to have killed a three-headed and six-eyed 
monster. It might be contended that the explanation is 
npt satisfactory, inasmuch as the head of Mriga is here 
supposed to be again conceived as a dog, while there is no 
authority in the Vedic works expressly describing Mriga as 
a dog. But if Orthros has become a dog in the Greek 
mythology, while it is a Mriga in the Vedas, I see no reason 
why Kerberos should not get his three heads from the 
Trishirshan of the Vedas. The difficulty is not at all a serious 
one. In bringing together the traditions of the three Aryan 
races after thousands of years, we must make some allow- 
ances, and be satisfied with a general similarity of the stories. 
The asterism of Mrigashiras and the dogs are so close, 
that one might be easily mistaken for the other, when all 
the knowledge of the original traditions was lost. It is thus 
that we can account for the fact that out of the three beings 
that were represented in this portion of the heavens, Bodra 
(the hunter), Mriga (bhe antelope), and Shv& (the dog), 
the Greeks retained in the sky only the hunter (Orion), and 
the dog (Kaon*, Ganis), with nothing to hunt, while the 
Hindus have not only forgotten, but condemned, the dog. 

The Parsis, it is true, have not mistaken the dog* but 
still as regards complexion, they have represented their 
dogs as possessing the colour which in the Bigveda is given. 

* The principal star in Ganis Minor is still called Procyon = Gk. 
Prokuon, Sk. Prashvan, the Foredog. It shews that the previous 
star was once called Kaon by the Greeks. If we count the Naksha- 
tras in the direction of the sun's annual course, Kuon comes first, 
and Prokaon afterwards. Of. Sanskrit Rddhd and AnuradM, of 
which, like Procyon, later writers have only retained AnwrMkl. 
Pkal'jum, Ashddhd, and Bhddrapadd, are similarly divided into 
Purod and Uttara,.t1a& preceding and the foregoing. 


to the antelope of the sun. Another objection that may be 
urged against this identification is that we are required to 
suppose Mrigashiras to be once the head of Prajllpati, and 
at another time that of Vritra. It must, however, "be re- 
membered that we do so on the express authority of the Rig- 
veda, and that besides it is quite natural to suppose that 
once the antelope's head was said to exist in the heavens., 
Vedic poets vied with each other in weaving legends out of 
it. As an illustration I refer to Rig. x. 86. 5, where the 
poet describes Vrishakapi's head as cut off, but soon after 
Yrishakapi is told that it was an illusion, and that in reality 
it was some one else whose head was so severed {verse 18). 
This clearly shows that it was a period when legends were 
stSl being formed out of the " antelope's head/* 

We can now explain how later writers evolved a myth 
out of Hamuchi's death. The story is given in the Tandy a 
Brahnlana (xii. 6. 8).* There we are told that Indra and 
Namuchi came to a settlement that the former should kill 
the latter, neither during day nor by night, nor by any 
weapon, whether dry or wet, Indra therefore killed him 
with the foam of the waters at the junction of day and night, 
when it had dawned, but yet the sun had not risen, It is 
probably this circumstance that has led Professor Max 
Miller to suppose that Orthros represents the gloom of the 
morning. Bnt the explanation does not account for the other 

* See also Taitt Br. L 7. 1. 7 ; Shat Br. xii. 7. 3. 3, Also 
the Pmrtes* Ramiyana iii. 30. 28 ; Mahabharata Udyoga p. is. 2. 
JProf. Bbomfielcl has collected all such passages in his article sa tltt 
cowtribmtbias to the Interpretation of tbe Veda in the Jonnarf irf 
American Oriental Society, Vol. XV., pp. 1 48-1 58. Tie kgjnfl f 
Hiraaya-Kashipu in the PnriQas appears to have beefli %a 

NamneliFs stoiy* 


incidents in the story. Was Namuchi or Vritra killed every 
morning by Indra ? Or was it only at the beginning of the 
rainy season ? Evidently the latter. We must then suppose 
that Namuchi was killed after dawn, bat before the actnal 
daybreak, at or during the monsoons. In other words, the 
junction of day and night in the later myths must be under- 
stood to mean a particular junction of day and night in the 
rains, or more definitely, the junction of the day and the 
night of the Gods the junction of the Pitriyana and the 
Devayana, the gates of which are said to be cleared np by 
Namuehi's. death in the passage from the Rigveda given 
above. The latter part of the legend is, however, still more 
poetical, and Prof. Max Muller's theory leaves it entirely 
unexplained. Indra is here said to have killed Narauclii 
with a weapon which was neither dry nor moist the watery 
froth. This is evidently based upon Rig. viii. 14. 13, where 
Indra is described as rt catting the head of Namuchi with. 
the foam of waters, " and the same incident is again referred 
to in Rig. x. 61. 8. Therefore, even if we reject later 
speculations with respect to "why foam or froth should have 
been used/' and decline to solve the question by assuming 1 a 
compact* between Indra and Namuchi, yet we have feo 
account for the fact that in the Rigveda itself Indra is said 
to have used the foamy weapon to destroy his enemy. What 

* Prof. Bloomfield has discussed this legend in a recent number 
of the Journal of the American Oriental Society (Vol. XV., Number 
II.), but he gives no explanation of the compact between ludra and 
Namuchi. In my opinion it is impossible to hold that the compact 
could have been the original basis of the legend. It is evidently a later 
invention to explain what were then deemed otherwise inexplicable 
incidents in the legend ; and until these incidents are' explained in a 
nastuial way, the legend cannot be said to be properly understood. 


could this foamy weapon be ? If Hamuchi was killed at the 
gates of the Devayana and his bead still lies there, tfee 
I watery foam conld be no- other than the broad belt of, the 
I -Milky Way which 'crossed the heavens at the same parti 
The blue vault of the heavens is often compared to an 
ocean in the later Sanskrit literature,* arid the stars -are said 
to be -the patches of foam upon its surface. -Thus in the 
Mahimtia Stotra, which is considered to be at IfeaBt 1 seven 
'or 'eigBi/ hundred years old, the author'- describes fv4^se f 17) 
the heavenly form of Rudra- (ii e. t Rmk& as /r'epreMhte'd in 
the sky), and tells ns that 'the strieafm of 'waters on his head 
9ias " the beauty of its foamy appearance enhanced by a 
ft,u&ber of stars/'t This is a description of the Ganges on 
*%he r hdad of the celestial form of Shiva, and the author of 
Ifahimna, who, in verse 22, refers to the story of RwJra 
piei-dmg Prajapati with an arrow, and says that the whole 
story 'is still illustrated in the skyjj evidently 'meant to 



\ * *Cf. Sahitya Darpana 10, where under 3TTlf[ we have 

^f ^^TR^H^nr^rrraBr crnr *f4f%T*f*rr: i 
t f^^rfr crnrnr^cr%^irjr^f%: 

Ir i 

Tie conception of Shiva embodied in this verse is really a gfa!nd 
one. The poet "asks his readers to imagine how great must Shiva 
be, the celestial* stream on whose head encircles the *TJiirYefse ! 
The Milky Way which girdles the celestial sphere camiot he 
better described. , -j > 

I AlsoCf. 


describe ;by. it- : tlie Milky /Way "which parses over the 
lieacL^f $ie star .of Eudra. Now if the poetic imagination 
of the -author of Mahimna can 'perceive foarti in the Milky 
Wayt>, r..I;-see no reason why the virgin, ira agination of the 
Vedic poets should not rise to that pitch. 13r. Haug^ 
speaking of the Vanant Yashta, observes that the constella- 
ting Variant),, by which the Parsi Dasturs understand the 
Hilty Way is said to stand directly 'over Hell,' and further; 
" the Dasturs are of opinion tha't this constfellaitidn : is "'"the* 
weapon (Vdzrd) which is constantly aimed* bj'Mithra'a^ 
the head of the Daevas, as stated in the Khurshed'Yasht."* 
Referring io 'the Khurshed Yashta we simply find that the- 
club (Vazra) of Mithra " was well struck down ic^on -the 4 
skulls of the Daevas.^f The information given to D'r, Haug 
may therefore be traditional among the Parsi Priests ; but 
whether traditional-' or otherwise, as : it comes from an 
independent source^ it 'is strong corroborative evidence to 
support the identification' of Indra*s foamy weapon*, with the 
stream of the Milky Way in 'the heavens. With the vernal 
equinox near the Dog-star, the' Mitky Way^ which then 
separated the region of gods 5 rom that of Yama, could well 
be said to be over Hell and- (t ^vell struck upon the heads 
of the Daevas.^ Namuchfs .legend can thus be simply 
and -naturally accounted ^for^ if we assign to the equinoxes 
the position^ which we have deduced from other passages in 
fchie V.edic -works. -I inay point but that we do not hereby 
account for tee original idea of Vritra. Tlat is evidently 
a stilt oldet 4egend;\ * But his existence at the gate of Hell 
and his decapitation by the foamy weapon the two chief 

* Dr. Haug's Essays on the Parsis, p* 271, note, 
t 6acred Books of the East Series, VoL, XXIIL, Zend Avesta, 
Part II, p. 87. 


elements in the later Yedic traditions are satisfactorily 
explained by placing, as originally proposed/ the vernal 

I equinox in the constellation of Orion, and identifying 
Namuchi alias Vritra with the constellation of Mrigashiras 

- or the antelope's head, situated just below the Milky Way. 

We have next to deal with the legends of the bold 
hunter, the terrible Eudra chasing the antelope. Several 
attributes in the Puranic mythology, e. g., his bearing the 
Ganges, in his matted hair, his fondness for the burning 
ground, and his appearance as Kirata or hunter, are all 
accounted for by placing Eudra just below the Milky 
Way or the celestial Ganges,* at the gates of the Ktri- 
yana and figured as a hunter. I have already alluded to 
the difficulty of identifying Rudra. But whether we take 
the star of Ardra or Sirius to represent the lord of cattle, 
the above attributes remain the same. But neither these 
legends, nor the story of Eudra chasing Prajapati, which, 
so far as it was necessary for our present purpose, has been 
already given, can help us, in a material degree, to solve 
the question under consideration. I wish, therefore, to deal 
here only with such traditions as point out to the position 
of Eudra in the course of the year. Rudra as the lord of the 
cattle and the presiding deity of storms, can be at once 
recognized and placed in the rainy season. There are, how- 
ever, other legends indicating time more definitely. In 
Rig. x. 192. 2, Samvatsara or the year is said to rise out of 
the ocean, the place where Vritra was killed (Rig. x. 68. 
12). Prajapati, as represented by Orion, may also be 
naturally supposed to commence the year when the vernal 
equinox was irt Orion. Rudra killed Prajapati, and as If 

\ \ 

* See Mahimna Stotra, verse 17, quoted supra. 


I have shown before, Prajapati, Samvatsara and Yajna 
were convertible terms. Rudra therefore killed Prajapati 
or Yajna at the beginning of the year ; and Yajna also 
meant sacrifice. Rudra was therefore naturally believed to 
have killed the sacrifice thus giving rise to the Puranic 
.legends of Rudra routing the sacrifice of Daksha. At the 
end of the Sauptika Parva in the Mahabharata * we are told 
that " Rudra pierced the heart of Yajna or Sacrifice with 
an arrow. Thus pierced the Sacrifice, with fire, fled away 
in the form of an antelope and having reached the sky, 
there shines in that form, followed by Rudra." Thus it was 
that Rudra acquired the title of Facrifice-breaJker. In the 
Tandya Brahmana vii. 2.1, the death of Prajapati is, however, 
spoken of as voluntary. In Taitt. Br. iii. 9. 22. 1, he is said 
to have assumed the form of Yajna and given himself up 
to the Devas to be sacrificed. The Devas killed Mm on 
their morning, and so every one should similarly perform the 
Ashvamedha sacrifice at the beginning of the year. One 
can now understand what the meaning of these stories is. 
They refer to the death of Praj&pati by Rudra at the begin- 
ning of the year ; and thus it was that Yajna, meaning- the 
year was sacrificed by means of Yajna or Prajapati, Rig. x. 90. 
16., where we are told that Gods sacrificed Yajna by Yajna, 
but this (human sacrifice) was an old (out of date) practice, 
may also be similarly interpreted. I cannot say which of 

\ * Maha. Saupt, 18, 13-14 : 

^r H 

Here the antelope is said to be pierced in the heart and not in. '* 
the head as in the Yedic works. It appears, therefore, that the 
whole antelope was considered to be in the heayens at this time. 


these legends is oldeiy : whether' tha| of Prajagat^crijicrag 
himself, or of Rudra killing him- at the beginning? of th$ yea,r. 
But whichever of these berthe , old$r. one it.dpes.npt Affect 
our present question. "Both of th am -indicate that Prajapati. 
once commenced -the year and. that' he .either, willingly? 
allowed himself to be sacrificed or .was killed ^y.JJudra at;. 
that time. As another indication of time, I may p'0 4 int5 .gut 
that the time prescribed for the- sacrifice of. SfatilagafV&jft, 
AshvalSyana Grihya Sutras, 4. 9. 2, is in Vasanta or/^ Shared; 
with the asfcerism of Ardra. The passage, 
stood ? means that the sacrifice should be r 
day in Vasanta or Sharad * whsn t the 

,, half ,; quarter or -new is near, the : .a 

hp star, over which Eudra .presides. : Bu,t. reappears, 
to, me that here we have a tradition t^hat-.the saprific^ 
was originally required to be pei-forpi^d %t^ the new or 
full moon in the vicinity of Ardra, in -V^santa or Sharad, 
thus indicating, that the vernal equinox .was near Ardra 
when the sacrifice was originally established. When th^ 
seasons receded Ardra new or full moon coukl^ not fall ini 
Vasanta or Sharad and therefore Ardrarnigh^t, g.^eryftrd 
mina, to, mean any night when the moon np^.-lh 
asterism of Ardra in Vasanta or Sharad- . jffpwav^r, .a tji^. 
point is not quite satisfactory I shall not h^.fj 1 ^^- 
only o%^ fact about L Rudra worthy of notice t i% c t t h.^t ,,h% 
seems td be described as followed by dogs or rather^as .theiii 
master (Vaj. San., 16. 27) .* This ?rnay shew that the if <*ie/ 
poets knew of the dogs hear the star of .Eudra. * 

. - ' - - - T^ - -J - - ; - - - ^ - , , 

* In the original there .are .salutations to .several' foipis of tb^, 
deity, but it would not t|e<jiwte safe to infer from it that Rudra wa% 
as a matter of certainty, followed by dogs/In t Tand. Br. xiv, J>?12, 
Shiva is described as Jf^gyti, while the passage m Taj. San. 
(16. 27) 


I have already alluded to 'the Pars! legends of the Chinvat 
Bridge and the dogs that keep it. There is, however, one 
TOr<earcumstance to which I wish here to refer. The star 
Tistrya'hs "been identified with.Sirius and the identification^ 
i^jadt absolutely correct, is at least sufficiently, so for general 
purposes. . But I think ; that the word itself has not been, yet 
satisfactorily explained. I propose to derive Tistrya from 
i-sivi which in- Sanscrit means three-stars, Tri-stri may 
r! be* corrupted. into Tistri, Tister. .Tister: is, therefore, 
stoeias -Kerb-eras or Tris'hiras and the f act 4hat Tistrya 
: 'is f -c&lled TSr 'or arrow in Modern Persian further confirms 
/'this derivation, -for the Aitareya Brahmana (iii. 83) calls it 
:: the' three-starred or tripartite arrow of Eudra in the sky. I 
'have in -the last chapter shown that if we continence with 
the summer solstice and regard Fravashinam a^ -the. first 
month of the year, Tistreye corresponds / to MargasMrsha. 
If Tister is understood etymolpgically to mean th^: belt of 
Orion this coincidence of the raonths-caiibe better accounted 
for. I am therefore of opinion that ' .Tistrya should not be 
r identified with Sirius, but with'- the belt of Orion. : We can 
~ -Ihen bbtter -understand why u the star should* have- Been 
of* as Tristryeni* probably indicating -more stars 
- iand also Pauryeni, -the first. The -Parsis have 

* As the word is understood,, at present it means ^pertaining to 
or belonging to Tristrya." But grammatically it may mean '* many 
stars or group of stars." I may here point out that if we identify 
Tistrya witl Shins the etymology is not explained, nor can we 
account for fie Modern Persian name Tir which again means an 
arrow. While if we identify Tristrya.with the three stars in the 
belt everything is satisfactbrily accountecT for. All the arguments 
based upon the "rain-producing" influence of the star are equally 
applicable in either case, since both the stars ( Sinus and Orion } 
rise at the same time* See Dr. Geiger's Civil of $ast Iran.,, Yel L y , 
pp. 14L-142. - , : .;,/ 


preserved another interesting relic of tlie asterism of 
Mrigashiras, but I reserve it for the next chapter. 

Starting with the supposition that the vernal equinox was 
in Orion, we have thus an easy and a simple explanation 
"by which the three principal deities in the Hindu mythology 
can be traced to and located in this part of heavens. 
Vishnu representing the happy times of Vasanta, Kudra 
presiding over storms and Prajapati, the deity of sacrifices 
beginning the year, were all combined in one place. It was 
here that Vishnu killed Varaha ( Rig. i. 61. 7) ; it was 
here that Indra killed Vritra, and it was here that Eudra 
chased Prajapati, iu the form of Yajna or that he sacrificed 
himself. r he celestial Ganges separating the upper and 
the nether world was also in the same quarters, and through 
|t Jay the path to Yama's region. In a word the Trinity of the 
i Hindu Pantheon was fully represented in the constellation 
of Orion, when the vernal equinox was there. Later writers 
describe this Trinity as represented by the three-headed 
Dattatreya, followed by the Vedas in the form of dogs; and 
after what has been said above, I think we can have no 
difficulty in identifying this personified Trinity with Orion 
having three stars in the head and closely followed by the 
dog (Canis) at its foot. It will be difficult to find another 
place int the heavens where all these elements are combined 
in such an interesting manner. 



Agrahflyqna = Agrayana in the older works Probable derivation of 
h'lyana The Agrayana, sacrifices Their number and nature Per- 
formed every half-year in Vasanta and Sharad Greek legends of 
Orion Their similarity to Vedio legends German traditions and 
festivities Stag and hind Twelve nights Dog-days All of which 
indicate the commencement of the year in Orion Dr. Kuan's explana- 
tion is insufficient The usual adjuncts of Orion His belt, staff and 
lion's skin The aiv-yaonghana, of Haoma in. the Avesta The yajno- 
pavtta of the Brahrnans Their sacred character probably borrowed 
from the belt of Orion or Yajna Use of mekhaU, ajina and danda, in 
the Upanayana ceremony Probably in imitation of the costume of 
Orion or Prajapati, the first of the Brahmans Derivation of Orion 
from Agrayana Its probability Phonetic difficulties Conclusion. 

IN the last chapter I have quoted an observation of 
Plutarch that the Greeks gave their own name to the 
constellation of Orion, and have there discussed some Vedic 
legends which corroborate Plutarch's remarks and indicate 
that the vernal equinox was in Orion at that time. In the 
present chapter I mean to examine other legends which go 
to shew that the constellation of Orion was known and 
figured before the Greeks, the Parsis, and the Indians 
separated from their common home, and that the legends 
or the traditions so preserved, and perhaps the name of the 
constellation, can be naturally and easily explained only on 
the snpposition that the vernal equinox was then near the 
asterism of Mrigashiras. 

I have already shown that Agr<ih3,yant 9 if not Agraliciyana, 

can be traced back to Panini's time, as the name of a 

Nakshatra, and that it is a mistake to derive it from the 

name of the full-moon day. We have now to see i we cam 



trace back the word still further. The word lidyana does 
not occur in the Bigveda, and it may "be doubted if the name 
Agralidyani was in use in the old Vedic days. Hay ana is., 
however, used in the Atharva Veda (viii. 2. 21 ; xi. 6. 17) 
and in the Brahinanas; and may be compared with Zend 
Zayano meaning winter, Panini (lii. 1. 148) derives hdyana 
from &$=== to go or abandon, after the analogy of gdyana, 
and gives two meanings, viz., the grain ' vrlhi ' and time. 
Whether we accept this derivation or not, it is at any 
rate clear that the word was used in Paninfs days, to denote 
a division of time and a kind of grain, and I think w 
can better account for both, these meanings of hdyana 
^ J ky connecting the word with ay ana and Agrayana or the 
JL VH half-yearly sacrifices. Dr. Geiger, speaking of the old Parsi 
cf calendar observes that ft probably the half-year was more 

employed in civil life than the complete year/'* Now 
whether the observation be entirely correct or not, we can, 
I think at any rate, assume that the division of the. year into 
two equal halves is an old one. I have already discussed 
the two-fold division of the year into Devaydna and Pitriydna 
and its coincidence with the passage of the sun to the north 
and the south of the equator. Ayana in the sense of such a 
division thus appears to be an old word and by prefixing h 
to It we may easily get hayana subsequently changed into 
, N fkdycwa like the words in the Prajnadi list, wherein this 
. Jword was not included as it was derived by Panini in a 
5 different way. The insertion and bmission of h when 

* Dr, Geiger's Civ. East, Iran., Vol; L, p. 152. Dr. Schrader 
makes a similar observation. "For all these reasons (most of 
which are philological) I believe we have the right to presuppose a4 
original division of the Imlo-Germanic year into two seasons.'* 
Preh. Ant, Ary. Peoples, Part IY., chap, vi, p. 302. 


followed by a vowel at the beginning of a word is not un- 
common even in these days,"* and there is nothing extra- 
ordinary if we derive hdyana from ay ana. Now by a 
natural process when we have two forms of a word or two 
derivatives of the same root they gradually come to be 
utilised for specific purposes, and so acquire distinct 
meanings. Sanskrit lexicographers class such words under 
Yogarudha, meaning thereby that etymology and conven- 
tion have each a share in determining their denotation. 
Hagana might thus come to exclusively denote a complete 
year, while ay ana continued to denote a half-year as 
before.-f When ayana thus became hayana, Agrayana, 
which all lexicologists derive from agra + ay ana, J would 
be changed into agra + liayana = AgraJiayana $ and when. 
Jiayana was changed to hdyana in a manner analogous to the 
words in the Prajnadi list (Pan. v. 4. 3&) as stated above, 
Agrahayana would be altered into Agrahdyana. We can 
thus account for the double forms Jiayana and hay ana , 
AgraJiayana and AgraJiayana which we find given, in 
Bohtlingk and Roth's and other lexicons, while if we ac- 

* Cf. The derivation of the word * history ' from e istory ' in 
Max Mailer's Lectures on the Science of Language, Vol. II,, p, 329". 

f Zend Zayano, denoting winter, probably preserves an older 
meaning, when hupam was used to denote the second of the two 
seasons (summer and winter) into which Dr. Schrader believes 
that the year was primevally divided. Some of the synonyms for 
the year in. Sanskrit originally denoted particular seasons, e. g., 
Parshd, Sharad. Samd and Hdyana may he similarly supposed to 
have been derived from the names of the half-year or ayana. 

J This derivation would give us Agrdyana instead of Agrayana 
and native grammarians obtain the second form from the first by 
the interchange of theinitial vowel with the following long a. 


cept Panini's derivation, hay aria will have to be either 
thrown out as incorrect or derived otherwise. In Amara 
ii. S. 52, hay ana occurs as a different reading for day ana in 
the sense of a vehicle and Bhanu Dikshita derives it from 
hay to go ; but we might as well ask it hay, ay, and i, all 
meaning to go, are not the different forms of the same root. 
As far as the form of the word is concerned we may there- 
fore derive hdyana from hay etna and the latter again from 
ayana and similarly Agrahdyana from Agrahayana and this 
again from Agrayana. 

I may> however, remark that the process which appears so 
simple according to the modern philological rules,, was not 
recognized by the native grammarians. There are good 
many words in Sanskrit which can be thus easily derived 
on the principle of the insertion and omission of h. Thus 
we have invakd and hinvdkd both meaning the stars on the 
top of Mrigashiras, and atta and hatta denoting a market- 
place. But native grammarians, including Panini, would 
not derive the words from each other, as we have done 
above in the case of ayana and hayana. Their method is 
to give two different roots for the two words ; thus we have 
two Vedic roots hinva and inva or hiv and iv, both mean- 
ing to go, to please, the one giving us hinvaltd and the 
other invaJcd. At and hat, an and han, ay and hay, i and 
hi are farther instances of the principle adopted by the 
native grammarians in such cases. Really speaking this is 
not solving the difficulty, but only shifting it a stage back- 
wards; for, if any explanation is necessary to account for 
the double forms like ayana and hayana, it is equally 
Tequirejr to explain why we should have the doublf roots 
like-oj and hay, both meaning to go. But it appears that 
the native grammarians, having traced the words to .their 


roots, did not push the matter further. With them ina is 
derived from i to go, ayana from ay to go, hayana from 
hay to go, and hayana from hd to go.* Whether and how 
far we can dispense with some of these roots is an impor- 
tant philological question, but it is not necessary for us to 
discuss it here. It does not much affect the point under 
discussion whether hdyana is derived from ayana, i. e., ay to 
go, or from hd to go as Panini has done. Etymologically 
both the words, ayana and hdyana, mean "going',' and 
when both carne to be used to denote a division of time, it 
is natural to suppose that they soon acquired special mean- 
ings. Thus while ayana continued to denote the half-year; 
hdyana, which was comparatively a later word, might have 
been exclusively used to denote the complete year, and as 
the beginning of the first ayana was also the beginning of 
the year, A(a)grayana would be naturally changed into 
A(a)grahd(a)ya,iia to express the beginning of the year. 
Whether we adopt Panini's derivation or the principle 
of modern philology we thus arrive at the same result, 
and so far as our present inquiry is concerned we can 
therefore suppose that the various words, which may be 
represented by A(a)gra(d)yana, or A(a)grahd(a)yana, are 
all transformations or derivations of agra + ayana = 

* This method sometimes fails, and native grammarians who 
are not now at liberty to coin new roots, have to resort to the 
Pnshodaradi list. For example, we have two forms ilvald and 
hilval'd as different readings for invaM in Amara i. 3. 23. Of 
these Undid can be derived from U 9 to sleep, though the root mean- 
ing" is ii@t suitable, but hilvald cannot be even so derived and 
Taranutha in his Vachaspatya would derive or rather obtain the 
initial h by Prishodaradi. Similarly cf. 
darldi! . .--. ' 


Now as regards tlie meaning it appears to me that ayana, 
at first denoted nothing more than the passage of the 
sun. Gradually it meant a division of time regulated by 
such passage. The Agrayana-ishtis thus appear to have 
originally meant the two half-yearly sacrifices performed 
on the first day of each ayana, which seems to be regarded 
somewhat like the new year's day at present. Gargya Nara- 
yana, in his commentary on Ashvalay ana's Shrauta Sutras 
<i. 2. 9. L) derives Agrayana from agra+ayana^ but interprets 
it to mean a sacrifice which is followed by eating (ayana), 
that is, which requires to be performed before the new 
harvest is used for domestic purposes. He thus takes 
ayana to mean eating, and as the Agrayaneslitis in later 
works like Manu (iv. 27) were described as u new-harvest 
sacrifices/' all commentators have adopted this explanation 
of the word. But it appears to me to be evidently of later 
origin and invented to account for the nature of the 
sacrifice when owing to the falling back of seasons the 
Agrayaneshtis came to be performed not at the beginning 
of each ayana as they should have been, but at wrong 
times. The necessity of such an explanation must haye 
been still more keenly felt, when instead of two half-yearly 
sacrifices, the Agrayanaishtis were performed thrice a yeai\ 
Ashvalayana, it is true, gives only two, one in Vasanta and 
the other in Sharad, the old beginnings of the Devayana 
and the Pitriy&na and the real commencement of the two 
ay anas. But he has mentioned three kinds of grain that 
maybe used, vrihi, shydmdka and yava ( i. 2. 9. 1, ) and 
his commentator Gargya Narayana observes that yava and 
shydmdka are to be used simultaneously in Sharad (i. 2* 9 
13). It appears, however, that the fact, that three kinds of 
grain were sanctioned for use, soon gave rise to three 


Agrayana-ishtis one in Yasanta with vrlhi ; the second 
in Varsha with shydmaka, and the third in Sharad with 
yava. But that it is a practice of later origin is evident 
from a passage in the Taittiriya Sanhita ( v. 1. 7. 3 ) 
which states that " twice is grain cooked for the year, " 
clearly meaning thereby thab there were only two 
Agrayana-ishtis in a year when the new harvest was first 
offered to gods. I am therefore of opinion that originally 
there were only two half-yearly sacrifices at the commence- ! i| 
ment of each ay ana, and as vrihi was used on the occasion 
of the first of these ishtis, the word ayana or kdyana 
naturally came to denote the grain SQ used, and that ayana 
in Agrayana originally meant not eating as the later 
writers have imagined, but a half-year as the word usually 
denotes. This way of deriving and explaining the word is 
not a new invention. For notwithstanding the fact 'that 
Agrayana and AgraJidyana are explained by Taranatha as 
referring to the sacrifice of grain and eating, yet he derives 
Agrayana, a word of the same group, from agra-\-ayana 
and explains it to mean that " the Uttarayana was in its 
front."* Even native scholars thus appear .to be aware 
of the fact that Agrayana could be or was derived from 
ayana meaning the Uttarayana. Indeed, we cannot other- 
wise account why the Agrayaneshtis were originally cele- 
brated at the beginning of Vasanta and the end of Var&ha 
as stated by Ashvalayana. The AgraJidyanl of Amara is 
thus traceable to Agrayani of the Vedic works ; and perhaps 
it was the initial long vowel in the latter that might have 
been retained in the later form. 

It may, however, be asked if there is any evidence to show 
* See Vachaspatya s. v. Agrayana. 


that Agrayana was used to denote a star in the Vedic 
works. That Amara, and long before him Panini, under- 
stood Agralidyanij if not Agrahayana, to mean the Naksha- 
tra o Mrigashiras is undoubted ; and I think we might 
fairly infer therefrom that the meaning given by these 
writers must have come down to them traditionally. Every 
ayana must begin with some Nakshatra, and it is quite 
natural to suppose that Agrayana must have gradually come 
to denote the star that rose with the first ay ana. But I 
have not been able to find out a passage where Agrayana is 
used in the Vedic works to expressly denote the constella- 
tion of Mrigashiras. I -may, however, refer to the Taittiriya 
Sanhita (vi. 4 11. 1.) wherein the vessels ( grahas ) used 
for sacrificial purposes are mentioned as beginning with 
Agrayana and considering the fact that two other vessels 
are named, as the words themselves denote, after the 
planets Shukra and Manthin 5 * we might suppose that Agra- 
yana came to be included in the list, not as the name of a 
deity, for it. was not such a name, but as denoting, the star 
which commenced the year, or the half-year. The word 
graha which in the sacrificial literature denotes vessel has 
been used in later astronomical works to denote the planets, 
the number of which, including the sun and the moon, is 
fixed at nine, the same as the number of the vessels used 
for sacrificial purpos es. It is not, therefore, improbable 
that Agrahdyanl or AgraJidyana of the later writers was a 
transformation of Agray an a, and that Mrigashiras, was so 
called in old times for sacrificial purposes. When the Agra 

* See infra Chap. VII. " In Taitt. San. iii. 1. 6. 3 the vessel is 
described as the vessel of Agrayana^ thus shewing that the vessel 
was named after Agrayana, which must therefore be either the 
name of a deity or of a Nakshatra. 



yaneslitis lost tlieir primary meaning, Agrayana or AgraJid- 
yana naturally came to be used more to denote the month 
when the sacrifice was performed than the Nakshatra at 
the beginning of the dyana, thus giving rise to the specu- 
lations previously discussed. But in whatever way we may 
explain the disappearance of Agrayana in the sense of 
Mrigashiras in the oldest Vedic works, the fact that in the 
days of Amara and long before hint of Panini Agrakdyani 
was used to denote the constellation of Orion remains 
unshaken, and we may safely infer therefrom that the^niean- 
ing giten by them was a traditional one. 

We have already seen how legends gathered round the 
(f antelope's head.'* It was the head of Prajapati wishing 
to violate his daughter, by whrch some understood the 
dawn, some the sky and some the star Aldebaran ( Ait. 
Br. iii, 33). Others built the sto^y of Namuchi upon the 
same which placed V|itra> at the doors of hell ; while a third 
class of legend-makers considered that the death of Prajapati 
was voluntary for the sacrificial purposes of the Devas, 
The following summary of the classical traditions about the 
death of Orion, taken from Dr. Smith's smaller Classical 
Dictionary, will show how strikingly similar they are to the 
old Vedic legends. 

" The cause of Orion's death is related Variously. 
<f According to some> Orion was carried off by Eos (Aurora), 
(e who had fallen in love with him; but as this was dis* 
f< pleasing to the gods, Artemis killed him with an arrow 
" in Ortygia.* According to others, he was beloved by 
" Artemis and Apollo t indignant at his sister's affection 

* Homer Od. T, 121. 4. See Gladstone's Time a,ud Place, of 
JSomer, p. 214. 
H t Q*. Fast v. 537. ' 


" for him, asserted that she was unable to hit with her 
" arrow a distant point which he showed her in the sea* 
" She thereupon took aim, the arrow hit its mark, but the 
<c mark was the head of Orion, who was swimming in the 
u sea. A third account, which Horace follows, states that he 
" offered violence to Artemis, and was killed by the god- 
f( dess with one of her arrows/ 3 

Thus love, arrow and decapitation which are the three 
principal elements in the Vedic legends, are all present in 
these traditions. There is another story which says that 
Orion was stung to death by a scorpion ; but this is 
evidently intended to represent the fact that the constella- 
tion of Orion sets when that of Scorpion rises in the east, 
and is therefore of later origin when the 20 diacal signs 
were adopted by the Greeks. 

There are other traditions which point out the position 
of Orion in the course of the year. The cosmical setting of 
the constellation was believed to be an indication of stormy 
weather and the constellation was called imbrifer or acquosus 
in the same way as the Shvd in the Vedas is said to corn* 
mence the year, while Shunasirau are invoked along with. 
Parjanya for rain. The German traditions are, however, 
more specific, and T take the following abstract of the same 
by Prof. Kuhn communicated to the late Dr. Rajendral&l 
Mitra and published by the latter in his " Indo- Aryans, 19 
Vol. II., pp. 300-302 :~ 

"Both in our ancient and modern popular traditions",, 
there is universally spoken of the Wild Hunter, who some- 
times appears under the name of "Wodan or Goden, and 
was, in heathenish times, the supreme god of the ancient 
German nations. This god coincides, both in character 
shape with the ancient Kudra of the Vedas (vide p t 


Now there is a class of traditions in which this ancient god 
is said to hunt a stag and shoot at it, just as Rudra in the 
Br^hmanas is represented as shooting at the rlski/a and 
rohit. The stag in German mythology, is the animal of the 
god Freyr, who like Prajapati, is a god of the sun, of ferti- 
lity, &c., so that the shot at that stag is to be compared 
with Rudra' s shooting at the Hs%a=Prajapati. I have 
farther endeavoured to show that some indications exist in 
tjie mediaeval penitentials of Germany and England, which 
give ns to understand that at the close of the old year and 
at the beginning of the new one ( we call that time ee die- 
ziwoiften" or the twelve clays, the dvddashdha of the Indians) 
there were mnmmeries performed by the country people, in 
which two persons seem to have heen the principal per- 
formers, the one of whom was disguised as a stag while the 
other was disguised as a hind. Both represented a scene, 
which mast have greatly interested and amused the people, 
but very much offended the clergy, by its sordid and hideous 
character; and from all the indications which are given in 
the text, communicated by me (pp. 108-180), we may safely 
suppose that the chief contents of this representation was 
the connection of a stag and a hind (or of an old woman), 
which was accompanied by the singing of unchaste songs. 
From English customs at the New-Year's Day, we may 
also infer that the hunter's shooting at this pair was even 
a few centuries ago, nay, is even now, not quite forgotten. 
Now as the time of the " twelve days * was with our ances- 
tors the holiest of the whole year, and the gods were 
believed to descend at that time from heaven, and to visit 
the abodes of men, we may firmly believe that this repre- 
sentation also was a scene of the life of the gods. I hope 
to have thus proved that the Brahmanical and the Germaa 


traditions are almost fully equal, and I have finally attempted 
to lay open the idea from which the ancient myth proceeded. 
According to my explanations, our common Indor-European 
ancestors believed that the sun and the day-light ( which 
was, so to say, personified under the image of various 
animals, as a cow or bull, a horse, a boar, a stag), was 
every day killed in the evening and yet reappeared almost 
unhurt, the next morning. Yet a (Jecay of his power was 
clearly visible in the time from midsummer to midwinter, 
in which latter time, in the more northern regions, he 
almost wholly disappears, ancjin northern Germany, Curing 
the time of the twelve days, is seldom 1 to b$ seen, the 
heavens being then usually covered all over with clouds. 
I have therefore supposed, it was formerly believed that 
the sun was then completely destroyed by a god, who was, 
both a god of night and winter as also of storm, Kudra= 
Wodan. The relics of the destroyed sun A th$y seem to have 
recognised in the brightest constellations of "the winter 
months, December and January, tfrat is, in Orion and the 
surrounding stars. But when they sa.w that they had been, 
deceived and the sun reappeared the myth gained th$ 
further development of the seed of Prajapati, from the 
remnants of which a new Aditya 9,3 well, as all bright an4 
shining gods were produced. I have further shown that 
both Greek astronomy and German tradition proved to be 
in an intimate relation with the Brahmanieal tradition ; for 
the former shows us, in almost the same place of the celestial 
sphere, a gigantic hunter (Mrigavyadha, Sirius; Orion/ the 
hunter Mrigaahiras) : whilst the latter has not yet forgotten 
that Saint Hiibertus, the stag-killei% who is nothing feui & 
representative of the god Wodan, who had, like Rudira, the 
power of healing all diseases ( the WmhaUcma * of the 


Vedas) and particularly possessed cures for rnad dogs which 
not only were his favourite companions; but were also in 
near connection with the hottest season of the year, when 
the declining of the sun begins, the socalled dag-days* 3 

Here is an equally striking coincidence between the 
German and the Vedic traditions. The mummeries were 
performed " at the close of the old year- and at the beginning 
of the new one," and the stag and the hunter had therefore 
something to do with it. Prof. Kuhn's explanation does not 
dear up this point satisfactorily, nor does it give any rea- 
son why the festivals were celebrated only during the v 
twelve days preceding the new year. As regards the decay 
of the sun's power it must have been observable during the 
whole season and does not therefore in any way account for 
the selection of 12 particular days. As for the duddashdha 
of the Indians, it is the period during which a person 
consecrates himself for a yearly sacrifice and so must natu- 
rally precede the commencement of the new year when the 
annual sacrifice commences, and I have previously shewn 
that it represents the difference between the lunar and the 
solar years ; in other words, they were whac we may now 
call the intercalary days added at the end of each year to 
keep the concurrence of the lunar and the solar measures 
of time. The German traditions therefore can be better 
accounted for, if we suppose that they are the reminiscences 
of a time when the stag and the hunter actually commenced 
the year. This also explains why the dog-days were consi- 
dered so important. When Sirius or the dog-star rose with 
the sun at the beginning of the year, the dog-days, or 
rather the days when the dog was not visible, were the new- 
year's days, and as such they were naturally invested with 
*ati importance which they never lost. I have already 


alluded to tlie passage in tlie Eigveda which states that the 
dog awakened the Ribhus, or the gods of the seasons, at the 
end of the year, and this appears to me to be the origin of 
what are still known as dog-days in the western countries. 
Owing to the precession of the equinoxes and by neglecting 
to maintain the correspondence of the seasons the days now 
fail during a period different from the one they did of old, 
but such differences we find in all cases where ancient rites 
or festivals are preserved. The feast of the manes, which 
the Parsis and the Hindus seem to have commenced to- 
gether when the summer solstice occurred in the month o^ 
Bhadrapada, now no. longer coincides with the summer sol- 
stice ; but for that reason we cannot say that it might not 
have occurred originally at the summer solstice, especially 
when the latter supposition is supported by other reliable 
evidence, and gives a better origin of the festival. I am 
not therefore disposed to accept Prof. Kuhn's explanation as 
satisfactory, and am of opinion that the German traditions 
are the reminiscences of a time when the vernal equinox was 
in. Orion, the hunter. We cannot otherwise "account why 
the mummeries and festivals should h,aye been celebrated 
during the twelve days at the end of the old and the begin* 
ning of the new year. 

It will, I think, be evident from this that the Greeks and 
Germans have preserved the memory of the days when the 
year commenced with the vernal equinox in Orion. I have 
previously shown that the Parsi primitive calendar, as fixed 
by Dr. Geiger, points to the same conclusion. The Parsis,, 
tlie Greeks, the Germans and the Indians therefore appear 
to have separated after these traditions were formed and 
after Orion was figured, and recognised as the Agrayana 
constellation. I do not think that any more traditional 


coincidences are necessary to establish tile Aryan, origin of 
the constellation of Orion, as well as its position at the ver- 
nal equinox in old days. I shall, however^ give one more 
coincidence which on account of its peculiar nature is alike 
interesting and important. 

In the Greek mythology Orion, after his death as above 
described, was placed among stars, " where he appears as a 
giant with a girdle, sWord, a lion's skin, and a club. >'"* Now, 
if as remarked by Plutarch, Orion is an original Greek 
name, we should find some traces of these various adjuncts 
of Orion or at least some of them in the old Iranian and 
Indian works. Do we so find them ? I think we do, only if 
we look for them with a little more attention and care, for 
the transformation is more specific and peculiarly out of the 
way in this case. In the Vedic works Soma is said to be 
the presiding deity of the asterism of Mrigashiras. Soma 
is Haoma with the Parsis* The 26th verse in the Haoma 
Yasht is as follows :- 

Frd te Mazddo "barat paitrvanim aivydonghanem 

steher-paesanghem mainyu-tdstem vanghuhim-daendm 


which has been thus rendered by Mr. Mills in his transla- 
tion of the Zend Avesta, Part III., in the Sacred Books of the 
Bast Series (p. 238) t " Forth has Mazda borne to thee, 
the star-bespangled girdle, the spirit-made, the ancient one* 
the Mazda- Yasnian Faith/' Dr. Haug takes faurvanwi 
in the original to mean " leading the Paurvas," which latter 
he believes to be the Persian name for the Pleiades, which 
is variously written paru, parvah, parvtn and parviz.-f Thia 
keen-sighted suggestion of Dr. Haug has been pronounced 

* See Smith's Dictionary of Classical Mythology* 
f Dr. Haug*s Essays on the Parsis, p. 182. 


by Mr. Mills as ec doubtful, and refuted by Vistasp Yashfc 
29, where Darmesteter renders a word probably akin as 
' the many/ " But excepting this difference of opinion all 
agree in holding this Yasht to be an ancient one, "a repro- 
duction of an Aryan original, 55 * and that the verse above 
given contains a description of the belt of Orion. Orion is 
Haorna, the Soma of the Indians which is its presiding 
deity in the Vedic works, and the above verse states that 
God has given a natural star-studded girdle to Haoma. 
This girdle is> therefore, no othr than the belt of Orion. 
The verse in the Haonla Tfasht, however suggests more 
.than it denotes. Both Haug and Mills have used the word 
* gir,dle ' in the translation. But whether we use 'girdle* 
or 'belt, 5 it hardly conveys the idea of the original aivyaon* 
ghanem. It is a striking instance of how in translations we 
sometimes lose the force of the original. Aivyaonghana is 
a Zend word for the kusti, or* the sacred thread of the 
Parsis, which they wear round their waist* The f girdle * 
or the ' belt ' of Orion is thus said to be his krtsti, and 
though we may have no more traces of the f belt* or the 
' club ' of Orion in the Parsi scriptures^ the above verse at 
once directs our attention to the place where we may expect 
^o find the traces of Orion^ belt in the Indian works* 
1 have before pointed out that Orion ot Mrigashiras is called 
Prajapati in the Vedic works, otherwise called Yajna. A 
belt or girdle or a piece of cloth round the waist of Orion or 
Yajna will therefore be naturally named after him as yajno* 
jpavHa, the upavita or the cloth of Yajna. The term, how* 
lever, now denotes the sacred thread of the Brahmans, and it 
may naturally be asked whether it owes its character, if ndb 

* See Sacred Books of the East Series, Vol. XXXL, 
art III., p. 238. . . . , 


the origin, to the belt of Orion. I think it does on the 
following grounds. 

The word yajnopavUa is derived by all native scholars 
from yafna + upavUa; but there is a difference of opinion 
as to whether we should understand the compound to 
mean an 'upavlta for yajna,' i. e.,for sacrificial purposes, or, 
whether it is the ' upavlta of yajnaS The former is not in- 
correct, but authority is in favour of the latter. Thus 
the Prayoga-writers quote a smriti to the effect that " the 
High Soul is termed yajna by the hotris* ; this is his upavita; 
therefore, it is yajnopavita." A mantra, which is recited on 
the occasion of wearing the sacred thread means, " I bind 
you with the upavlta of yajna ;"f while the first half of the 
general formula with which a Brahman always puts on his 
sacred thread is as follows : 

The mantra is not to be found in any of the existing 
Sanhitas, but is given in the Brahmopanishad and by 
Baudhayana. This verse is strikingly similar to the verse 
quoted above from the Haoma Yasht, It says, " yajno- 
famta is high and sacred ; it was born with Prajapati, of 
old/* The word purastat corresponds with pa-urvanim in 
the Avesta verse and thus decides the question raised by 


t See TMaatha's Yashaspatya s. v. upavUa ; and Saukhya- 
yana Gnhyasiitra, ii. 2. 3, where the mantra is given, as follows : 
Hf^ I ^Rff^l ceJT ^h T Pft^*iiHH^IHT i I a ^he Paraskara 

Gphyasutra, ii. 2. 11 ,. both these mantras, ^rr^Tcf <rc*f &c., and 
c., are given. , , ; ,,^, f 



Dr. Haug, while sahaja* born with the limbs of Prajapat-i, 
conveys the same meaning as mainyu-tastem. The coinci- 
dence between these verses cannot be accidental, and it 
appears to me that the sacred thread must be derived from 
the belt of Orion. UpavUa, from ve to weave, literally 
means a piece of cloth and not a thread, f It appears, there- 
fore, that a cloth worn round the waist was the primitive 
form of yajnopavUa, and that the idea of sacredness was 
introduced by the theory that it was to be a symbolic 
representation of Prajapati's waist-cloth or belt. In the 
Taittiriya Sanhita ( ii. 5. 11. 1.) nivita, prdchinavita, and 
upavita, three words which at present denote the position 
of the sacred thread on the body of a Brahman,, are defined, 
but the MimansakasJ understand them to apply not to the 
sacred thread, as we now wear it, but to a piece of cloth 
or deer-skin, which everyone must use at the time of 
sacrificing. It appears, therefore, that iu the oldest times 
the Brahtnans wore a piece of cloth or deer-skin and not 
a thread. This conclusion is further strengthened by the 
fact, that according to the ritual given in the Sutras, no 
sacred thread is mentioned in the description of the 
ceremony of Upanayana; while the investiture with the 


com.* on Brahmopanishad (MS.). 
* t Cf. Medhatithi on Manu, ii. 44. 

J Cf. Jaiminiya-nyaya-maia-vistara, iii. 4. 1. 3T?f 

( Taitt. Arn., ii. 1 ) ft^ *TpT^r*U ^**T ^ ft^ffaf 
JHH?. I Taitt. Arn. ii. 1 is the only passage in the V^rc works 
which fully describes the positions R?frcf &c., and it expressly 
mentions 3TCT and s?f%?r, but not *rr. 

See Taranatha's Yachaspatya s. v. upawta. Also A$hvalaya"gtt 
Grihya Sutra j. 19, 8-10-12, where ajina, mehhald, ^id danda 
are alone mentioned.^ ',.-... 

VI'.] . ORION AND HIS BELT. 147 r 

thread is looted upon at present as tlie principal part of that? 
ceremony, We hare still retained a memory of this old 
practice in the performance of obsequies and at the time of 
performing sacrifices, when a piece of cloth is worn in addition 
to the sacred thread. Devala* says that oat of the three 
sacred threads to be worn, one is a substitute for the upper 
garment, thus clearly indicating what the old practice was. 
Bat this is not the place to go into these details. It is 
enough for our purpose to notice that yajnopavita originally,- 
meant a piece of cloth, and that in the times of the smriti- 
writers, it came to be symbolically represented by the sacred 
thread, thrice twisted and thrice folded. Here is, however, 
another difficulty which must be here noticed. The Parsis 
wear their sacred thread round the waist, while the Br&hmacs 
usually wear it over the left shoulder and across the body, 
leaving the right arm free (i. e. 3 upavita). The Parsis may" 
thus be said to wear their sacred thread after the manner 
of Orion; but in the ca-se of the Brahmans, it may be ques- 
tioned if their manner of wearing the thread corresponds 
to the position of Orion's belt. From the passage in- 
the Taittiriya Sanhita referred to above, it will, however, 
be seen that nivtta (and not upavita), is the position of the 
thread there prescribed for all human actions, or, in other 
words, for doing all ordinary business of life. Nivita has been 
defined by all later writers to mean, the position of the 
sacrecl thread passing around the neck, over both the 
shoulders and dropping down in front. A reference to 
Kumarila BHatta's Tantra Vartika (iii. 4. 2.), will, however, 
show that nivtta, also meant " tying round the waist/' and 
Kumarila observes that " tying round the waist is the most 


convenient position for all kinds of work/'* Ahandagiri 
and Govindananda in their commentaries on the ShankanVs 
Bhashya on the Brahmasutras ( iii. 4, ]9. ), give the same 
explanation, from which it appears that the Brahmans, 
like the Parsis, once wore the thread around the waist, thus 
literally girding up their loins when they had to do any work. 
The sacred thread of the Parsis and the Brahmans thus 
seems to be a symbolical representation of Prajapati's girdle 
or Orion's belt in every respect. The various stages, by which 
the original piece of cloth round the waist dwindled into 
a thread, are interesting and instructive from a ceremonial 
point of view, but not being r-elevant to the present inquiry, 
I do not mention them here. 

* As the passage is important as a record of now obsolete 
practice I give it here in the original 

i %r^RjT?p 
T 3jTJffr% i 

The word ^Rfrf in this passage indicates that the writer had a 
^flf text in his mind. Madhavain his commentary on theParAshara 
Smriti (Cal. Ed., p. 450) quotes Kutyayana and Bevala as follows : 

I think these verses clearly indicate that the thread must be 
worn below the breast and above the navel, and goiqg round the 
whole waist. As the practice has long since been obsolete, the 
verses have been much misunderstood by later writers. The 
author of the ^R3f*fcrK does, however, clearly state that there 
are two ways of wearing the thread, first over the shoulder as 
described in the Taitt. Am. ii. 1 ; and ( 3Tf r in the original ) 
second as given in the above texts of K&tylyana an& 
This view has also been adopted by the authop of fje 


Bat the sacred thread is not the only trace of Orion's 
dress that we have retained. A reference to the Upanayana 
ceremonial will show that we have preserved. |belt, staff, 
skin, : and all. Every boy, who is the subject of this ceremony, 
has to wear a mekhald or grass cord round his waist, and 
we still put three knots to this cord just over the navel, as 
it were, to represent the three stars in the belt of Orion."* 
In the Vajasaaeyi Sanhita 4. 10, we are told that the knot 
of the melthald; when it is worn for sacrificial purposes, is to 
be tied with the mantra, " you are the knot of Sorna/'-f* 
which Mahidhara explains as " a knot dear to So ma ;' f but 
which remembering that we have a similar verse in the 
Haoma Yasht, may be naturally interpreted to mean the knot 
of Soma, the presiding deity over the constellation of Orion- 
Then every boy whose tip may ana, or the thread-ceremony 
as it is popularly understood, is performed, must carry with 
him a stick of the paldsha or the fig-tree and the same 
passage in the Vajasaneyi Sanhita says that for sacrificial 
purposes the stick (danda) is to be taken in hand by the 
Mantra, " wood ! be erect and protect me from sin. till 
the end of this yajna." Here again Mahidhara interprets 
yajna to mean sacrifice for which the staff is takea up. 

* In the Prayoga works we have (and we still do so) : 

In the S&nkhy&yana Grihya Sutra ii. 2. ,2, we are told that the 
knots of the mekkald may be one, three or five, and the commen- 
tator adds that the knots should be equal in number to one's 
pravaras. The author of the Sanskara Kaustubha quotes a smriti 
to the same effect. But the explanation is unsuited to the first 
case, viz., of one knot, and I am inclined to take it to be a later 



But I tliink here also we may trace a reference to 
Prajnpati alias Tajna. The third accompaniment of a 
newly initiated boy is the deer-skin. Theoretically it is 
necessary that he should be fully clothed in a deer-skin, but 
practically we now attach a small piece of deer- skin to a 
silk-thread and wear this thread along with the yajnopavita. 
HeJchald, ajina, and danda ( the girdle, the skin and the 
staff ) are thus the three distinguishing marks of a newly 
initiated boy ; and what could they mean, except that the 
boy is made to assume the dress of Prajapati as far as 
possible. To become a Br&hman is to imitate Prajapati, the 
first of the Brahmans,- Prajapati assumed the form of a 
deer, so the boy is clothed in a deer-skin ; Prajapati has a 
girdle round his waist (the belt of "Orion), so has the boy 
his mekhald with three knots over the navel ; and lastly, 
Prajapati has a staff, and so the boy must have it too.* 

* Dr. Schrader in his Preh, Ant. Ary. Peop., Part iv., Chap. 
viii., concludes that the primitive dress consisted of a piece of 
woolen or linen cloth thrown round the shoulders like a mantle, 
and a girdle. The history of yajnopavita, the way of wearing it 
as described in Taitt. Am. ii. 1., and Orion's dress, as conceived by 
the Greeks, point to the same conclusion. I have already alluded 
to the difficulty of explaining how upavtia, which literally means a 
cloth, came to denote a thread. If yajnopavtta be taken to have 
originally meant yajna and upavita, and yajna be further supposed 
to kave once denoted a girdle this difficulty is removed. Av. yasto 
Gk. zostos, Lith.justas, meaning "girded" point to an original ro0t 
jos, Av. yangk y from which Gk. &dnu, Av. aiv-y donghana may be de- 
rived (See Ficks* Indo-Germ. Wort,)- If we suppose that the root 
appeared as yaj in Sanskrit and derive yajna from it, like Gk. zo 
we may take yajna to mean a girdle and translate 

Wrfr*:(JftbaLUpa.5.)by "how can a Brahman be withouta girdle and 
-a cloth ? JI If this suggestion be correct, then yajndyavSfa must be 




Thus in their Dpanayan ceremony the Brahmans have fully 
preserved the original characteristic of the dress of Praja- 
pati or Orion. The Brahman latu (boy) does not, however, 
carry a sword as Orion is supposed to do, and the skin used 
by the boy is deer's and not lion's. I cannot account for the 
first of these differences except on the ground that it might 
be a later addition to the equipment of Orion, the hunter. 
Bat the second might be traced to a mistake similar to that 
committed in the case of the seven rikshas. The word 
Mriga in the Rigveda, means according to Sayana both a 
lion and a deer, and I have already referred to the doubts 
entertained by modern scholars as to the animal really de- 
noted by it. Mrigdjma is therefore likely to be mistaken 
for lion's skin. There is thus an almost complete coinci- 
dence of form between Orion as figured by the Greeks and 
the boy whose upanayana is recently performed, and who is 
thus made to dress after the manner of Prajapati. I do not 
mean to say that a piece of cloth was not worn round the 
waist before the constellation of Orion was so conceived ; on 
the contrary, it is more natural to suppose tbat the amount 

taken to have meant nothing more than a mantle and a girdle in 
primitive times and that the primitive people invested Orion with 
a dress similar to their own. "When Orion came to be looked upon, 
as a celestial representation of Prajapati, Orion's dress must have 
attained the sacred character which we find preserved in the sacred 
thread of the Parsis and the Brahmans. I, however, know of no 
passage in the Vedic literature where yajna is used in the sense of 
a girdle, and hence the above suggestion must be considered as 
very doubtful. But it may be here mentioned that in Maratht we 
use the wor&jdnve to denote the sacred thread. This word is 
evidently derived from. Sk. yajna, Prakritajanflo. Perhaps we 
have retained only the first word of the long compound yajnojpavUa. 


people invested Orion with their own dross. But the coinci- 
dence of details above given does, in my opinion, fully 
establish the fact that the sacred character of a batu's 
dress was derived from what the ancient priests conceived 
to be the dress of Prajapati. With these coincidences of 
details, still preserved, it is impossible to deny that the 
configuration of the constellation of Orion, is of Aryan 
origin and that the Hellenic, the Iranian and the Indian 
Aryas must have lived together when these traditions and 
legends were formed. 

And now it may be asked that if the Eastern and the 
Western legends and traditions of Orion are so strikingly 
similar, if not identical, if the dress and the form of the 
constellation are shewn to have been the same amongst the 
different sections of the Aryan race, and if the constella- 
tions at the feet and in front of Orion Canis Major and 
Canis Minor, Tuon and Prokuon,* Shvan and Prashvan, the 
Dog and the Foredog are Aryan both in name and tradi- 
tions ; in short, if the figure, the costume, the attendants and 
the history of Orion are already recognised as Aryan, is it 
not highly probable that the name, Orion, should itself be 
a transformation or corruption of an ancient Aryan word ? 
prion is an old Greek name. Homer in the fifth book of 
Odyssey speaks of the bold Orion and the traditional coin- 
cidences, mentioned above, fully establish the probability 
of Plutarch's statement thab the word is not borrowed from 
a noti- Aryan source. Two of the three names, mentioned 
by Plutarch Canis (Kuon) and Ursa (ArJdos) have agaia 
been phonetically identified with Sanskrit shvan and rikshag, 
and we may, therefore, legitimately expect to find Orion 
similarly traced back to an Aryan original. The task, ILOW- 

* See note on page 119 




everts not so easy as it appears to be at the first sight-. 
The Greek mythology does not give us any help in the 
solution of this question- It tells us that a hunter by name 
Orion was transformed after his death into this constellation 
which consequently came to be called after him. Bnfc this 
is surely no satisfactory explanation. Who is the hunter 
that was so transformed ? There are many mythological 
proper names in Greek which can be traced back to their 
Aryan originals, and why should Orion be not similarly 
derived ? The story obviously points to the Vedic legends 
of Rudra, who is said to be still chasing Praj&pati in the 
heavens. The Vedic legend has fully preserved all the 
three elements in the story the hunter Rudra, the dog 
and the antelope's head, while the Greeks appear to have 
retained only the hunter and the dog with nothing to hunt ! 
QBut that does not, preclude us from discovering the 
identity of these legends, and the question is whether 
we can suggest a Sanskrit word which -will give us Orion 
according to the already established phonetic rules. I know 
of no name of Rudra from which Onon can be so derived; 
But if we look to the names of the constellation of 
Mrigashiras, we may, I think, in the absence of any better 
snggestion, provisionally derive Orion from Sanskrit 
Agrayana the original of Agrahdyana. The initial long d 
in Sanskrit may be represented by 'omega in Greek as in Sk; 
dma, Gt. dmoSi Sk. dshu, Gr. okus, and the last word ayana 
may become ion in Greek. It is not, however, so easy to 
account for the dropping of g before r in the body of th4 
word. Comparison of Sk. grdvan with Gk. laos and of Ski 
ghrdna with Gk. ris, rinos, shews that the change may takf 
place initially, but scholars whom I have consulted think 
that there is no instance in which it takes place 
20 ' "" " """" ' "" '"'" 


between Greek and Sanskrit, though such changes are not 
rare between other languages as in Old Irish &r, Cymric 
aer, which K. Bruginann J derives from * agra. Also com- 
pare Gk. dakru, Goth, tagr, Old Irish der, English tear ; 
Latin exagvnen, examen, 0. Ir. dm 9 from the root aj. I do 

r not feel myself competent to decide the question, and hence 
must remain content with simply throwing out the sugges- 
tion for what it is worth. I have shewn that traditional 
coincidences clearly establish the possibility of the Aryan 
origin of Orion, and if I have not hit upon the correct word 
that does not affect my argument. My case does not, in 

{fact, rest on phonetic coincidences. I rely principally upon 
certain statements in the Vedic works, which indicate that 
the vernal equinox was once in Orion, and I wanted to shew 
and I think I have shewn it that there is sufficient 
evidence in the Greek and Parsi legends to corroborate the 
statement in the Vedic works about the Phalguni-full- 
moon being once the first night of the year. "We can now 
give a reasonable explanation of how Fravarshinam came 
to be the first month in the primitive Parsi calendar and 
why Dathusho should have been dedicated to Din (creator) ,f 

J Comp. Gram., Vol. I. Arts, 518, 523. Prof. Max Muller 
extends the rule to Greek and Latin, see his Lectures on the 
Science of Language, Vol. II., p. 309, where several other instances 
are given. For a full statement ^of the phonetic difficulties in 
identifying Gk. Orion with Sk. Agrayana, see App. to this essay. 

t By the bye it may be here remarked that we can perhaps 
better account for the names Ahuramazda and Ahriman on 
the theory that the vernal equinox was then in Orion, the 
winter solstice in Uttara Bhadrapada and the summer solstice 
in Uttara Phalgirat The presiding deities of the last two Nakshatras 
are respectively Ahir-Budhnya and Aryaman. According to the 
A vesta belief, which assigns the south to the gods and the north to 
the Daevas, Ahir-Budhnya, as the regent of the southernmost . , 



The mummeries and festivals amongst the Germans can 
also be more satisfactorily accounted for, while above all, 
the form, the dress and the traditions of Orion may be now 
better traced and understood. I have already in the 
previous chapter shown that even the Vedic legends, espe- 
cially those in the later works, can be simply and naturally 
explained on the assumption we have made regarding the 
position of the equinoxes in the days of the Rigveda. (The 
hypothesis on which so many facts, legends, and traditions 
can be so naturally explained, may, in the absence of a 
better theory, be fairly accepted as correct without more 
proof^But in the present case we can go still further and 
adduce even direct evidence, or express Vedic texts, in its 

point, would come to be regarded as the supreme, ruler of the gods, 
while A.ryaman would be the king of the evil spirits. Therefore we 
may suppose that the names Ahura-Mazda and Ahriman, if not 
actually derived from these words, were, at least modelled after 
them. Amongst the names of the Yedic deities Ahir-Budhnya is 
the only word, both the component members of which, are declined 
as in Ahura-Mazda. Spent-Mainyus and Anghra-Mainyus is a 
distinct pair by itself; and besides the difficulty of deriving Ahri- 
man from Aiigkra-Mainyus, there seems to be no reason why 
Ahriman, if so derived, should be contrasted with Ahnra-Mazda- 
($e<?Phil. Mazd. Eelig. by Casartelli, trans, by F. J, Dastur Jamasp 
Asa, 71, 72, pp. 54>-6_) Pa,rsi mythology has another deity 
named Airyamaa, and as this word is derived from Sanskrit Arya- 
man, it tnay be objected that same word cannot be said to have also 
given the name for the evil spirit. I do not think that the objection 
is well founded. Of. Andra(Sk. Indra) and Verethraghna ( Sk. 
Vritrahan ) both of which are the names of the same deity in 
Sanskrit, but one of which has become an evil spirit in the Avesta. 
But I cannot fully discuss the subject in a note, and not being 
pertinent to my case, I cannot also do more than merely record 
here an explanation that may possibly be suggested. 


support. In the chapter on the Krittikas, I have drawn 
attention to the remarks of Prof. Max Miiller who objected 
to the conclusion based entirely on the Vedanga Jyotisha 
on the ground that no allusion to the position of the Kritti- 
kas \vas to be found in the Vedic hymns. We can now 
account for this silence; for how can the hymns, which 
appear to be sung when the sun was in Orion at the begin- 
ning of the year, contain any allusion to the period when 
the vernal equinox fell in the Krittikas ? This could have 
been easily perceived if, instead of confining to the contro- 
versy about the position of the Krittikas and endeavouring 
to find out if some clue to the date of the Veda could bo 
obtained from the determination of the original number and 
source of the Nakshatras, scholars had pushed their inquiries 
further back and examined the Vedic hymns in the same 
critical spirit. It would not have been difficult in that 
case to discover the real meaning of the Vedic verse which 
states that "the dog awakened the Ribhus at the end of the 
year/' I have in a previous chapter already referred to the 
verses in the Rigveda regarding the position of Jama's dogs 
and the death of Namuchi. These passages,, as well as the 
description of Vriha or the dog-star risiag before the sun 
after , crossing the eternal waters; the terminus of the 
Devayana (Rig. i. 105. 11.), sufficiently indicate the position 
of the equinoxes in those days. In the next chapter I 
propose, to discuss and examine two other important passages 
from the Rigveda, which directly bear out the statement in. 
the Taittirya Sanhita with which we have started, vi%>, that 
the Phalgunf fall-moon commenced the year at the winter 
solstice in days previous to those of the Taittirya SanMt4 
and the Brahrnanas. 



Knowledge of astronomy in Veclic times The seasons and the year The 
ay anas The zodiacal belt or rita Observation of a total eclipse of 
the sun in the Rigveda Knowledge of the planets Shukra and 
Manthin Venus and Vena, Shukra and Kupris The legend of. the 
Ribhus Their identification with the Ritas or the seasons of thet 
year Their sleep or rest in Agohya's (sun's) house for 32 intercalary 
days Said to be awakened by a dog (Rig. i. 161. 13) at the end of the 
year Indicates the commencement of the year with the dog-star : 
Nature and character of Vrishakapi His identification with the sun 
at the autumnal equinox The hymn of Vrishakapi in the Bigveda 
x. 86 Its meaning discussed verse by verse Cessation and com- 
mencement of sacrifices on the appearance and disappearance of 
Vrishfi kapi in the form of a Mriga IndrAn! cuts off his head and seta 
a dog at his ear Orion (Mrigashiras) and Canis Meaning of nedfajas 
in the Vedic literature When Vrishakapi enters -the house of Indra, 
his Mrlga becomes invisible (Rig. x. 86. 22.) Points to the vernal 
equinox in Orion or Mriga Leading incidents in the story stated and 

IT is said that we cannot suppose that the Vedic "bards 
were acquainted even with the simplest motions of heavenly 
bodies. -The statement, however, is too general and vague 
to be criticised and examined. If it is intended to be 
understood in the sense that the complex machinery of 
observation which the modern astronomers possess and the 
results which they have obtained thereby were unknown in 
early days, then I think there cannot be two opinions on. 
that point. But if by it is meant that the Vedic poets were 
ignorant of every thing except the sun and the dawn, 
ignorant of the Nakshatras, ignorant of months, ay ana**, 
years and so on/ then there is no authority or support for 
such a supposition in the Rigveda. On the contrary., ,w* 


find that some of the Nakshatras are specifically named, 
such as Arjtini and Ayhd in Rig. x. 85. 18, while the same 
hymn speaks generally of the Nakshatras, and the motions 
of the moon and the sun as causing the seasons. In Rig- 
i. 164 we have again several references to the seasons, the 
year and the number of days contained in it (verse 48) and 
according to Yaska, perhaps to the ay anas (Nirukta 7. 24), 
I have in a previous chapter referred to the passages 
in the Rigveda, which mention the Devayana and tlie 
Pitriyana, the old names of the ay anas beginning with the 
vernal equinox ; and there is, therefore, no objection to 
understand the above verse (i. J64. 48) as alluding to the 
black or the Pitriyana. The intercalary month is mention- 
ed in Rig. i. 25. 8, while in i. 24. 8 Varuna is said to have 
constructed a broad path for the sun, which appears 
evidently to refer to the Zodiacal belt. I am further in- 
clined to think that the path of rita (Rig. i. 41. 4) which 
is mentioned several times in the Eigveda, where the 
Adifcyas are said to be placed (s. 85. 1), and wherein 
Sarama, discovered the cows of Indra (v. 45. 7, 8) refers to 
the same broad belt of the Zodiac which the luminaries, as 
observed by the Vedic bards, never transgressed. It was 
so to speak their 'right' way, and therefore called rita, 
which though literally derived from ri, to go, soon came to 
mean lie 'right* path, the circle of which exists for ever, or 
rather exists and exists (varvarti) in the vault of the heavens 
(Rig. i. 164. U). Prof. Ludwig goes further and holds 
that the Eigveda mentions the inclination of the ecliptic 
with the equator (i. 110. 2) and the axis of the earth (x. 86. 
4). It is now generally admitted that the seven rifohas 
were also known and named at this time. The mention of 
a hundred physicians in Rig. i. 24. 9 may again be taken 


to represent tlie ftsfcerism of Shala-ljliibhalz or Shata- 
tdrakd, presided over by Varnna according to the later 
lists of the Nakshatras in the Taittirrya Brahrnana. 
The fortieth hymn in the fifth Mandala of the Bigveda is 
still more important in this connection. It shows that an 
eclipse of the sun was then first observed with any preten- 
sions to accuracy by the sage Atri.* It is thus that I 
understand the last verse in the hymn which, after describ- 
ing the eclipse, says, cc Atri alone knew him (the sun) none 
else could/ 3 This observation of the solar eclipse is noticed 
in the Sankhyayana (24, 8) and also in the Tandya Brah- 
mana (iv. 5. 2; 6. 14) 3 in the former of which it is said to 
have occurred three days previous to the Visliuvdn (the 
autumnal equinox). The observation thus appears to have 
attracted considerable attention in those days. It seems 
to have been a total eclipse of the SUB, and the stars became 
visible during the time, for I so interpret the- expression, 
bhuvandni adidJiayuJt in verse 5. In verse 6 we are told 

* Prof. Ludwig has tried to deduce the date of the hymn from 
this circumstance. But the attempt is a failure as shewn by Prof. 
Whitney (see the Proceedings of the American Oriental Society, 
Yol. XIII., pp. 17-22 ). As the eclipses recur in the same order after 
a certain period, we cannot use such facts for chronological purposes 
without knowing the geographical position of the place where the 
eclipse occurred, and even then the conclusion will be correct only if it 
can be shown on independent grounds that such a phenomenon did not 
occur at that place during several centuries before or after the date 
we determine. I, therefore, simply use the hymn for the purpose 
of showing that an eclipse of the sun was observed in those day sin such 
a way as to leave a record behind. It would be difficult to deduce any 
other reliable conclusion from it even upon the assumption, not known 
and hence not used by Prof. Ludwig, that the vernal equinox was 
then in Orion and that the eclipse occurred three days before the 
autumnal equinox as described in the Brahnumas. I cannot, however, 
accept the suggestion that the hymn may be understood as referring 
to the obscuration of the sun by clouds. 


4;hat "Atri knew (the eclipsed sun) by turiya Irahma^ and 
Sayana interprets the last two words to mean ic the fourth 
verse or mantra/ 3 But the verse wherein these words occur 
as itself the sixth, and Sayana has to explain that by " fourth" 
is to be understood the '* fourth, if we count from the sixth, 
i. e., the tenth verse ! " The explanation may be good 
from the ritualistic point of view, but it appears to me to be 
t]uite unsatisfactory otherwise. I could rather interpret 
tur'h/ena brahmand to mean " by means of turiya" Turiya 
is mentioned in modern astronomical works as a name for 
an instrument called quadrant (SiddMnta Shiromani xi. 15), 
and though we may not suppose the same instrument to 
have existed ia the old Vedic days, yet there seems to be 
too objection to hold thafc it may have meant some instru- 
fetent of observation. The word brahma is no doubt used 
to denote a mantra, but it may also mean knowledge or the 
means of the acquiring such knowledge. In Rig. ii. 2. 7 
Sayana has himself interpreted braJtma to mean some ct act 
or action; " and I see no reason why we should not under- 
stand the phrase turiyena brahmand in the above hymn to 
mean "by the action of turiya" or, in other words, "by means 
of turiya" and thus give to the whole hymn a simple and 
natural appearance, rather than endeavour to interpret it 
after the manner of the Red Indians, who believed that 
Columbus averted the calamity of the eclipse by prayers. 
The peasants of the Vedic times, some scholars might 
argue, cannot be" considered to be more civilized than the 
"Bed Indians ; but in so arguing they forget the fact that 
tiier e must be a Columbus, who would, by his superior 
Capacity, inspire the feelingp of awe and reverence for him. 
When the |mrd% therefore, tell us that Atri knew, of the 
eclipse by turiya br^Jm, we can new easily &ee what it means, 

VII. ] 



Sayana's explanation, as I have above observed, may be good 
from the ritualistic standpoint ; but we cannot, for other 
purposes, accept an interpretation which makes the ' fourth* 
to mean the ' tenth J verse of the hymn 1 Thus understood 
the hymn clearly indicates that at the time when the 
observation was taken the Vedic priests were tolerably well 
acquainted with the elementary astronomical facts. It is, 
however, suggested that the planets were unknown in these 
days. I am unable to accept even this statement. It is 
impossible to suppose that the Vedic poets, who constantly 
watched and observed the various Nakshatras in the Zodiac, 
should not have noticed planets like Venus, Jupiter, or 
Saturn^ which outshine many of the Nakshatras in brilliancy. 
The periodical appearance of Venus in the west and the 
east* and especially its rising only to a certain altitude followed 
by its regress, are facts too striking to remain unnoticed even 
by the superficial observers of the heavens. But we mast 
not go on mere probabilities. The hymns of the Rigveda are 
before us, and though probabilities may serve the purpose 
of determining the direction of our search, yet if we cannot 
find any reference to the planets in the Vedic works them- 
selves, we must give up the notion that they were known to 
the poets of these hymns. There is no question that planets 
were known in the days of the Brahmanas. In the Taittiriya 
Br&hmana (iiu 1. 1* 5) we are told that Brihaspati (Jupiter) 
was first born * near the asterism of Tishya, and to this 
day the conjunction of Tishya and Jupiter is considered as 
highly auspicious in the astrological works. We have, how- 
ever, to look for any allusion to the planets in the Rigveda 

. I This reminds us of 
Big. iv. 50. 4, where similar wording occurs, thus : 



itself. The mention of the five bulls in Rig. i. 105. 10 may 
not b considered as sufficiently explicit to denote the five 
planets; * but what shall we say to the mention of Shukra 
and Manthin together in Rig. iii. 32. 2 and ix. 46. 4 ? They 
seem to be evident references to the vessels called Shukra 
and Manthin used in sacrifices and have been so interpreted' 
fey the commentators. Bat as I have before observed, the 
vessels in the sacrifice themselves appear to have derived 
their names from the heavenly bodies and deities known at 
the time. It is generally conceded that the sacrificial 
arrangements more or less represent the motions o the 
sanf &&d the chief events of the year. In other words, 
tte yearly sacrifice is nothing but a symbolical representation 
0r -rather Imitation of the sun's yearly course. If so, it is 
t suppose that some of the sacrificial vessels at* 
were named after the Nakshatras and the planets. 
la the Taitfciriya Sanhitd iii. 1.6. 3. the vessels are spoken 
of as ' the vessel of Shukra, ' 'the vessels of Manthin, ' and 
so on, which indicates that Shukra and Manthin were nofc 
wad as adjectives of the vessels. The only other explana- 
tion is to suppose that Shukra Manthin, Agrayana, &c., were" 
the names 01 Son*a juice, and that the vessels used for 
IE its various capacities, were described- 
Shukra, &a There is, however, no authority' 
for holding: that Soma really ha* 

so different mpaeitie ; and I therefore 

Bictotioa of Sfenfera and Manthin, a^ 

ty note 7 m p^t 11^ I hold that 'the 

t >> , 


applied to vessels, in the Kigveda is a clear indieatidn o 
the planets being then discovered. There is, how.ever, in 
my opinion, a more explicit reference to a planet in th0 
Bigveda which does not seem to have yet been noticed. 
In the tenth Mandala we have a hymn (123) dedicated tp 
Vena^ which according to Yaska denotes a deity of the 
*tniddier region. Yaska (Nirukta 10. 38) derives the word 
from ven ' to love/ f to desire, ' and explains it as denoting, 
as his commentator Durgacharya says, " loved by all; n * 
while the hymn itself contains such expressions as the " son 
of the sun/' ** on the top of rita," " comes out of the ocean 
like a wave, " t &c., which have been variously interpreted 
by the commentators. But from all these facts I think we 
have herein the original Aryan name of Venus. The word, 
'or rather the meaning I have here proposed, is entirely 
lost in the Sanskrit literature, but considering the fact 

that the Latins named the planet as Venus, while the word 

t ** ei szeB&tF'-- 
cannot be satisfactorily derived from any Latin root/} 

titeera can be BO objection to identify Venus with the Vena 
(nom. sin, Venas) in the Vedic works. In the Latin- 
mythology Venus is the goddess of love, and this we can now 
easily account for, as the name of the Vedic deity is derived 
.from a. root which means " to desire, " " to love." I may 
again point out that the hymn of Vena in the Rigveda y is 

* See Mahidhara on Vaj* San. 7. 16. Some consider that ihe 
root is Fin and not Few.. . . . 

f This reminds one of the tradition of Aphrodite who, in Sreek 
mythology, is said to be sprung from the foam of the sea. 

J In Dr. White's Latin-English Dictionary the word is derived 
/rom Sanskrit van to love; but if it is to be derived from a Sanskrit 
-root why not derive it from yin or ve to desire or;1oye, and so 
connect it with Veyia of tbre ^lUgveda. ., ,1 J :., t 


used In sacrifices at the time, when the priest' takes tip the 
Tessel Shnkra in the sacrificial ceremonies.* Katyayana, 
indeed, mentions the optional use of the hymn for taking 
up the vessel of M anthin.t But that does not much alter 
the position, for, when the meaning of the word was utterly 
forgotten the hymn might come to be used for a different 
purpose in addition to the previous one. The fact, that 
the Vena hymn was used in taking up the Shnkra vessel 
is, therefore, an important indication of its old meaning, and 
when we find the name actually preserved till now indicat- 
ing the planet Venus, and that this name cannot be 
satisfactorily derived in any other way, we might fairly 
infer that Vena of the Rigveda is Venus of the Latin 
Mythology* As regards the change of gender we need not 
eonslder it to be a serious objection inasmuch as not only 
Venus, but also the moon has changed in gender in its 
passage to Europe. As a further proof of the statement 
that the planets, or at any rate Shukra, was discovered and 
named in the primitive period, I refer to the Greek word 
Kupris (Latin Cypris) which means Venus. The word cau 
fee easily identified with Sanskrit Shukra which, according 
to the well-established phonetic rules, becomes Kupro in 
Qmk, lie initial sh being changed to Jc, as in Sk. shvcm, 
HL trnrn and the medial Jsr to pr by labialisation, cf. Gk. 

* See Bar^chlrya on Nirukta 10. 89. ^ *TfrOT*T *J*nr. 
f Tfce Sttzts of Katyayana hearing on this point are as follows 


( 6. n-13) *j* fMfr * r ff qrtfe tinf 

%f fS|% !4ftNH*3? tft r . Thus he first lays down that the 
Slmtei vessel should be taken by reciting the hymn ^ 
( VaJ. Sa. 7. 12 ) or aecordiiig to some the hymn BT 
J-T^'8.Z-.K,.Kg. ;X .,-l). He then observe* 
Jailer hymn is used in taking ^p the Manthin vessel. 




jnriamai, Sk. &r-ttmt, I purchase. As Venus was supposed 
to be a feminine deity in Europe Kupros was naturally 
changed into Kupis. Thus, both the Latin and the Greek 
names of the deity may be traced back to the Vedic Vena 
and Shukra, and we may therefore hold that the planet 
was discovered and named before these races separated* I 
know that European lexicologists derive Kupris from 
Kupros the Greek name of the island of Cyprus where Venus 
was said to be much worshipped and that Cyprus again is 
supposed to have received its name from the trees, cypresses, 
in which it abounds ! But the explanation, which gives no 
derivation for the name of the tree, seems to me to be 
quite unsatisfactory. If Aphrodite was known to the Greeks 
in the primitive times it is more natural to derive the name 
of the island from the name of the deity. In course of time 
this original connection between the name of the deity and 
that of the island may have been forgotten, and Greek writers 
regarded Eupris as born in Cyprus. But we must receive 
these derivations of Greek mythological proper names with 
great caution as most of them have been suggested at a 
time when comparative Philology and comparative Mytho- 
logy were unknown. Latin cuprum meaning f copper * is 
again said to be derived from Cyprus (Gk. Eupras), but it 
does not affect our argument, for whatever be the reason for 
giving the name to the island, once it wa^ named Cyprus or 
Kupros, many other words may be derived from it without 
any reference to the reasons for which the island was so called. 

Some of the reasons given above may be doubtful, trot on 

the whole I am inclined to hold that the Vedic Eishis 

were not as ignorant of the broad astronomical facts 

*as they are sometimes represented to be. They mem fe 

have watched and observed &e sun and 


their yearly course noted the bearing of their motions on 
the division of time, fixed the length of the solar year and 
endeavoured to make the lunar correspond with it. The 
Nakshatras and thsir rising and setting also appear to have 
been, duly observed. It was perceived that the sun and the 
moon and such of the planets as they had discovered never 
travelled out of a certain belt in the heavens, called rita ; 
while the eclipses of the sun and the moon also received due 
attention and notice. Men, who were acquainted with these 
facts, would naturally be able to fix the beginning of tlie 
months and the year by the stars that Tose at the time, and 
though we cannot suppose the Vedic bards to have been in 
tie possession of any accurate astronomical instruments, yet 
it was not difficult for them to decide roughly by simple 
observation when .the day and the night were equal, OF 
when the sun turned to the north, either from the solstitial 
or from the equinoctial point. The knowledge implied by 
these observations may appear to be too much for a Vedic 
poet in the opinion of those who have formed their notions 
of primitive humanity from the accounts of savages in 
Africa or the Islands of the Pacific. But as observed before 
/we must give up these a priori notions of primitive 
Jhumanity. in the face of evidence supplied by the h^mns of 
the Bigveda. It is on this evidence that we have to form 
our judgment of the primitive Aryan civilization, and if that 
evidence is found conflicting with our prepossessions, thj& 
latter must be given up. In what follows I shall therefore 
assume the capacity of a Vedic bard to make the simple 
observations above mentioned. 

. We shall now take up the verse in the Bigve^ referred 
ftOiSeveyal times, previously, the verse, which, declares tha,fc 
a dog awakened tk^Ribljuf; at the epd w of ^ y$$r. (Big, 


i. 161. 13) ; and the first question that arises in this con- 
nection is, who are the Ribhus ? Referring to Nirukha 
(11. J5 and 16) we find that native scholars consider that 
the three Ribhus Ribhu, Vibhvan and Vaja were the 
sons of Sudhanvan and that having rendered wonderous 
services to the Gods they gained divine honors and a share 
in the sacrifice and immortality.* But even Yaska does 
,not seem to be satisfied with this explanation. There are 
several hymns in the 'Kigveda wherein the deeds of the 
Ribhus are described (Rig. iv. 33-37; i. 20. 110. Ill and 
161), and in most of them the Ribhus are spoken of aa 
working in close connection with, the year (samvatsam or 
samvatsara}. Thus in the Rigveda i. 110. 4 they are said 
to have commenced work at the end of the year, and in iv* 
33. 4. they are described as engaged, for the whole year 
(samvatsam ), in reviving the cows (the rays of the sun) < 
The Ribhus are further mentioned as resting in the house 
of Agohya, the " unconcealable " sun-god for twelve days 
at the end of their course (Rig. iv. 33. 7). In Ait. Br. iii. 30 
they are described as sun's neighbours or pupils (ante vdsds)* 
while in. Rig. iv. 51. 6 their work is said to be done by the 
dawn. , Ya^ska therefore considers that the Ribhus also re- 
presented, the rays of the sun, and in this he is followed by 
Sayana. But the explanation does not account for the number. 
of the Ribhus who are said to be three brothers. We must 
therefore go a step further and hold that the Ribhus did 
BO t merely represent the rays of the sun generally, but the 
three seasons, as connected with them, as several European, 
scholars hare suggested.f In the Rigveda iv. 34* 2, the 

* Also compare Brihad-devata iii. 81. 88 ; p. 82, Cal. Ed., where 
the same story is given. 

" f See Kaegi's Rigveda, p. 37, and note 12-7 on page- H&i . Parti-s 
oularly see Lad-wig's Rig.''i&*-p$ 187-.9:. . . ' , _.. . ,.,..,/ _,,_ ^ 


Ribhus" are told to rejoice with the seasons (Ritus) and this 
supports the latter view. In Eig. i. 15. 10, Dravinoda is 
said to be the fourth companion of the seasons and the 
Shatapatha Brahmana (xiv. 1. 1. 28) expressly states that 
there are three seasons. It is therefore generally believed 
that this was the old division of the year, and that the 
number of the seasons was increased as the Aryas travelled 
further from their original home.* The three Eibhus, 
representing the three seasons, may thus be said to be 
engaged, throughout the whole year, in doing wonders for 
the gods and received as guests in the house of Agohya at 
the end of their course. " Here they spend twelve days in 
enjoyment; then the course begins anew, and anew the 
earth brings forth fruit, the streams flow; plants cover the 
heights, and waters the depths."f And now comes the* 
verse (Rig. i. 16 L 13) on which I rely: 

Here the Eibhus, awakened from their sleep and rest for 
twelve days, ask " Agohya ! Who is it that awakened us ?"J 
The goat (the sun) replies that it is the " hound. " Sayana 
understands shvdnam to mean ' wind, ' but there is na 
authority for it and the meaning is perfectly unnatural. In 
fact Sayana may be said to have failed to interpret the 

* Kaegi's Bigveda, p. 116, note 68, where he quotes Zimmer to 
the same effect. 

f This is in substance a translation of Rig., i. 161. 11 and iv^ 
33. 4, See Kaegi's Rigveda, p, 37. * 

J Idam in the first line is not the object of abtibudhat as Sayand, 
and Mr. S. P. Pandit suppose. It should be taken either in appo- 
sition with tat> or as an adverb meaning * now/ * here,' &c. 




verse correctly. Ludwig and Grassmann both translate ifc by 
f hound/ but neither of them explains what it signifies. 
There is again some difference of opinion as to whether 
the word samvatsare should be taken with bodhayitdram 
or with vyakhyata. But whichsoever construction we adopt 
the meaning remains the same, since it is the same thing 
if the Ribhns are said to be awakened at the end of the 
year and then commenced their conrse, or they awakened 
and then looked up at the beginning of the new year, or, in 
other words, commenced their new-year's course. Practically, 
therefore, all agree in holding that the awakening of the 
Ribhns here referred to is their awakening at the end of 
the year, after they have enjoyed sound sleep and rest in 
the house of Agohya for twelve ( intercalary) days, and the 
only question that remains is, who is the hound or the dog 
that awakens them ? We have seen that the Ribhus were 
the genii of the seasons and that as companions of the sun 
they worked wonders during the whole course of the year. 
But as it was a lunar year, 12 days were intercalated at the 
end of each year to make it correspond with the solar year. 
These 12 days belonged neither to the old nor to the new 
year, and the Ribhus were therefore naturally believed to 
suspend work during this neutral period and spend it in 
rest and enjoyment in the house of Agohya. When the 
whole legend has thus a chronological signification it is 
natural to hold that the hound, here alluded to, must be 
some constellation in the heavens, and if so, after what has 
been said in the previous chapters about it, what could it 
be except Canis Major or the Dog-star V The end of the 
year here referred to is evidently the end of the three 
seasons, represented by the three Ribhus, and we must, 
therefore, take it to mean the end of the equinoctial yeat 



or tlie beginning of Vasanta, the first of tlie seasons. Durga- 
charya in Ms commentary on Nirukta 11. 16 explains the 
phrase samvatsare ( in Rig. i. 110. 4 ) in the same way. 
As I have already discussed the subject before,* I do not 
here repeat the grounds on which I hold that the year, in 
primitive times, commenced with the vernal equinox. 
Prof. Ludwig has made a happy suggestion that dbhogaya, 
which the Ribhus are said to desire (Rig i. 110. 2) before 
they commence their career and reach the house of the 
sun, should be interpreted in its ordinary sense to mean, 
the bend or the inclination of the ecliptic with the equator. 
Our investigation, based upon, independent facts, leads us 
to the same conclusion. In short, the whole story of the 
Ribhus, as we find it recorded in the Rigveda, directly 
establishes the fact that at the time when this legend was 
formed the year commenced with the vernal equinox in 
Canis Major or the Dog-star. It is highly improbable, if 
not impossible, to give any other reasonable interpretation 
to the verse in question, whether we understand the Ribhus 
to mean the three seasons of the year or the rays of the sun 
as Yaska and Suyana have done. With the vernal equinox 
near the Dog-star, the winter solstice would fall on the full- 
moon in Phalguna and Mrigashiras would head the list of 
the Nakshatras. Our interpretation of the verse in question 
is, therefore, fully warranted by the traditions about the 
ancient year-beginnings given in the Taittirfya Sanhita and 
the Br&hmanas. 

Let us now examine the too much and too long mis- 
understood or rather not-understood hymn of Vrish&kapi 
in the tenth Mandala of the Rigveda. As there is only one 
hymn in the Rigveda which gives the story, it is not so 

* See myyra, Chapter II. 

'< ' - L "rf. 1 1 '"// > i 
. ' ' 




easy, as in the case of the Bibhus, to determine the nature 
of the deity, and hence various conjectures have been made 
by scholars as to its origin, character and meaning. The 
deities appear both in the masculine and in the feminine 
form, Vrishakapi and Vrishakapyi. Amara* considers 
that Vrishakapi means either Vishnu or Shiva, and Vrisha- 
kapayi either Lakshm! or Gaurf. In the Brihad-devatH 
Vrishakapi is said to represent the setting sun, and 
Vrishakapayi the gloaming-f Yaska (12. 27) would derive 
the word so as to mean the sun who shakes (the world) 
with his rays, and his commentator observes that the god 
showers mist or dew and shakes the animate world. 
Modern speculations about the derivation and the meaning 
of the name may be found in Bhanu Dikshita's commentary 
on Amara (iii. 3. 130). Prof. Mas Miiller, in one place,J 
observes that ' e it is difficult, on seeing the name of Vrisha- 
kapi, not to think of JBrika/paeos, an Orphic name of Profo- 
gonos and synonymous, with Phanes, HeUos, Priapos, 
Dionysos," but, says he, "the original conception of Vrisha- 

* Amara iii. 3. 130 and 156. 

Brihat-Devata ii. 9. and 10 : 


And, again further on in ii. 69 and 70. 

$ Lectures on the Science of Language, Vol. II., p. 539. 


kapi (Vrishan, bull, irrigator ; Kapi, ape, tremulous) is not 
much clearer than that of EriJcapaeos." However, if the 
comparison be correct, we may, I think, take it as confirm, 
ing the identification of Vrishakapi with the sun proposed 
by several scholars, native and European. In fact, there 
seems to be a general agreement that Vrishakapi represents 
the sun in one form or the other. But this alone does not 
account for all the incidents recorded in the hymn. I 
would, therefore, further suggest that VrisMkapi be under- 
stood as representing the sun at the autumnal equinox, 
when he may be rightly said to shake off the rains, inas- 
| much as the equinox falls at the end of the rainy season. 

I have previously shown that the conception of Vishnu and 
{ Shiva can be traced to the Vedic Vishnu and Rudra, and 

these latter may be taken as the types or tho embodiments 
of the mild and terrible aspects of nature at the vernal and 
5 the autumnal equinox. If VrisMkapi in later mythology 

i has therefore come to denote Vishnu and Shiva, according 

to Amara, the meanings are consistont with the supposition 
that in the Vedas Vrish&kapi represents the sun at the 
I equinoxes. In the hymn itself, Vri&hakapi is said to have 

I assumed the form of a yellow antelope whose head IndrSni 

is described to have cut off. This circumstance serves to 
guide us in at once fixing the position of Vrishakapi in the 
| heavens. It is the same antelope's head that has given rise to 

I so many myths. When the position of V ri^hdkapi is thus fixed, 

| it would not be difficult to understand the various incidents 

| described in the hymn. But without further anticipating 

|. what I have to say in the explanation of the hymn, I 

now proceed to examine the hymn itself, We shall then 
see whether the assumption which we have made regarding * 
the character. &d attributes of VrihHkapi gives u* & 


simple, natural, and above all, intelligible explanation 
of the story given in the hymn, which, as explained at 
present, is nothing but a bundle of disconnected, if not 
mutually inconsistent, statements. I shall first quote the 
original hymn. 


TO 1 Ik ^Rfff f 







I? t ft 


1 1 1 ^ If 



There are twenty-three verses in the hymn ; and of these 
3, 4, 5 and 20, 21 and 22 have a direct bearing on the 
question we are discussing. But to understand these verses 
properly, it is necessary to discuss most of the other verses 
in the hymn, and I shall therefore examine the hymn verse 
by verse. I have already remarked that the hymn is one of 
those which have not yet been properly understood. Some 
of the verses have been explained by Yaska, but he has 
nownere tried to give us the bearing of the whole story 
described in the hymn. S&yana's commentary is very often 
simply verbal, and in many places he too is not certain 
about the meaning, while the AnukramanJ has been several 
times disregarded by S&yana himself. On the other hand, 
Ludwig, Grassmann, and several other European scholars 
have tried in their own way to explain the legend or the 
story embodied in the hymn, and the latest attempt of the 
kind is that of Piscel and Geldner in their Vedic studies, 
Vol. VII., Part I.* These scholars hold that the hymn 

* I am indebted for this information to Dr. B. G. Bhand&rkar, 
who kindly undertook to explain to me the views of German 
scholars on this point. 


narrates a legend current in old days. In other words, they 
take it, and I think rightly, to be a historic hymn. Bat the 
question, what does the legend signify, or how did it 
originate, still remains unsolved. Piscel and G-eldner 
understand the hymn to mean that Vrishakapi went down 
to the south and again returned to the house of Indra. 
But even then the bearing of the legend is but imperfectly 
explained. The occurrence of such words as ddsa, drya and 
parshu in the hymn have led some to suppose that the hymn 
records the story of a struggle between the Aryan and the 
non- Aryan races. But the hypothesis hardly explains the 
various incidents in the story, and the legend may therefore 
be said to be but still imperfectly understood. Under such 
circumstances any suggestion which explains the hymn 
better is at least entitled to a hearing. It is admitted that 
the hymn is a dialogue between Indra, Indram, and Vrisha- 
kapi, a son of Indra as they call him.* But there is a great 
divergence of opinion in assigning different verses to their 
deities. I shall examine these points while discussing the 

Katyayana in his Sarvanukramani says f 

Upon this the Vedartha-dipika by Shadguru- 

shishya has, 

Nr: 1 The verses of the 

hymn are then distributed amongst the speakers as follows : 
Indra, 1, 8, 11, 12, 14, 19, 20, 21, 22 ; Indram, 2, 3, 4, 5, 6, 9, 10, 
15, 16, 17, 18 ; and Yrishakapi, 7, 13, 23. The same distribution is 
given in the Brihad-devat& by Shaunaka. Piscel and Geldner in- 
troduce Vrishakapayi in the dialogue and distribute the verses 
somewhat differently thus, Indra, 1, 3, 8, 12, 14, 19, 20 1 ; 
JndrAnt 2, 4, 5, 6, 9, 16, 21 ; Vrishakapi 7, 10 13 ; and VrisM- 
kapAyi, 11, 15, '17, 18. Verses 22 and 22 are supposed to* be 
addressed by a stranger, the narrator. 


VERSE 1. The verse has been differently interpreted 
"by different writers. Yaska (13. 4.) interprets it as 
referring to the rays of the sun, which ( the rays ) deem 
themselves perfectly independent of the luminary by which 
they were sent out. Say ana ascribes it to Indra who savs 
" that sacrificers, allowed by me to sacrifice fco Vrishakapi, 
have disregarded me, but are praising the lord Vrishakapi 
who is delighted, as my friend, in the sacrifices, where 
plenty of Soma is used; [but notwithstanding] this Indra 
is superior to all.' 3 Madhava Bhatta, whom. Sayana mentions 
with respect, however thinks otherwise. He considers 
that the verse is addressed by Tndrani to Indra, when she 
.perceived that the sacrificers have ceased to sacrifice on 
account of the oblations being spoiled by an animal repre- 
senting Vrishakapi. He would, therefore, thus interpret 
the verse. Says Indrani, " In places of plenty where lord 
Vrishakapi revels, sacrificers have given up sacrificing and 
disregarded Indra. My friend Indra is superior to all/'" 

When the very first verse is thus interpreted in three 
different ways,* one can easily attribute the difference to an 
imperfect perception of the bearing of the whole hymn. 
To me Madhava Bhatta alone appears to have taken into 
consideration the verses that follow. Thus the fifth verse 
of the hymn states that the things of Indrani were spoilt by 
Vrishakapi in the form of an animal, and consequently he 
was beheaded. I should, however, like to refer to verse 21, 
wherein Vrish&kapi is told that when he appears again, 
sacrifices would be performed. This evidently implies that 
they were stopped before and were to be commenced again 
on tibfe tfe-t{^earance of Vrishikapi. The first verse there- 
fore b$ ; interpreted to mean, that "the sacrifices are 
The root sn/, with vi may mean either to 


abandon or to allow, but the former is its natural meaning, 
and when verse 21 in the same hymn can be easily explain- 
ed by taking the former meaning of srij with vi, it 
would be straining the words if we put a different inter- 
pretation on them. I am, therefore, disposed to interpret 
the verse after the manner of Madhava Bhatta, except the 
last sentence. 

Bnt why should sacrifices be stopped ? What has 
Vrishakapi to do with them ? These are very important 
questions, and I am sure that had they been properly 
answered, there would have been no difficulty in interpreting 
the hymn. In verse 3 we are told that Vrishakapi? 
spoken of in this hymn, has the form of a yellow antelope. 
) 1 In verse 5 Indrani is prepared to cut off his head, because 
tf**' he offended her, and in the preceding verse ( 4th ) a dog 
is said to be let loose upon him. These facts an antelope 
,! with the head cut off, and a dog closely following him are 
- *" quite sufficient for the purposes of identification. They 
shew that the whole story is based upon the " antelope's 
head " we have previously discussed ; and bad Yaska and 
Sayana known that there is a constellation called dog iu the 
heavens by the side of Mrigashiras, I feel certain that they 
would not have hesitated to recognize in Vrishakapi, the 
sun as jtpES8fi&ted by the constellation of Orion. But all 
traces of the dog, as a constellation, having been lost in the 
Sanskrit literature, neither Yaska nor Sayaiia could find 
any clue to the true meaning of the hymn. This is not, 
however, the only place where Yaska has been obliged to 
invent extraordinary interpretations. Not knowing that 
the dog represente^La,,star, he has proposed (Mrakta 5. 20), 
tha 'Vrika, sTiouTd be understood to mean '"the 'moon,** 
while usually it means a wolf or a wild dog, and it appears to 


me that a similar mistake has been also committed here* 
Comparative Mythology and Greek Astronomy have, how- 
ever, thrown further light on the subject, and we mnst now 
try to interpret the hymn accordingly. Vrishakapi must, 
therefore, be taken, to represent the sun in Orion. 
,; Bat even supposing . that Vrishakapi thus represents the 
sun in Orion, why should the sacrifices be stopped on his 
account? The identification of Vrishakapi with Orion at 
once furnishes us with a solution of this question. We 
have already seen that the dog is said to commence the new 
year in Rig. i. 161. 13; and since Canis and Orion are close 
to each other, Orion may also be said to have commenced 
the year. The Devayana, therefore, extended in those days 
from the heliacal to the acronycal rising of Orion ; that is> 
when Orion rose with the sun, it was the vernal equinox, 
the beginning of the Devayaua, and six months after, when 
it rose at the beginning of night, it was the autumnal 
equinox, the end of the Devayana. Now all Deva- 
cereinoaies and sacrifices could be begun and performed only 
during the Devayana, * or, as we find it in later traditions* 
only in the Uttar&yana. The acronycal rising of Orion was 
thus a signal to stop such ceremonies, and oblations could 
properly be said to have been spoilt by the appearance of 
this constellation at the beginning of night.t But above 
all the burden of tlie song " Indr# is utiara of all,'? 
becomes specially appropriate in this case. The word utiara 

* Jaimiai Mim. Dar, vi. 8. 23, auct other authorities cited in 
Chap. II. 

t If Vrit-ra is correctly identified with the constellation of 
Mrigashiras, we may on the same theory also explain why he is 
called Makhasyu in Rig. x. 73. 7. The appearance of Mriga, at 
the beginning of nisht, indicated the commencement of the Dakshi* 
nay ana when sacrifices were stopped. Vritra alias Mriga might 
thus come to be regarded as a destroyer of the sacrifices- 



* obes not here mean superior, but f6 upper'* implying that 
Indra is in the upper or the northern portion of the universe, 
though the sun or Vrishakapi may go down. I would 
therefore translate the first verse thus : " Where my friend 
Vrishakapi rejoiced in the wealth of the Aryans, they gave 
up sacrificing and did not respect Indra. Indra is (how- 
ever) in the upper (i. e., northern) part of the universe/'* 

VERSE 2.. Indra is here reproached for following up 
Vrishakapi, though he has offended Iridrani. Says sli-e to 
Indra : 6< Indra ! (how is it that) you run down fast after 
Vrishakapi and do not go anywhere else to drink sorna. 
Indra is, &c. ?) 

The word pard in this verse seems to denote the region 
where Vrishakapi has gone. Pardvat is often said to be 
the place in the distant or lower portion of the sphere, and 
is thus contrasted with arv&vat (Rig- viii. 13. 15). In Big. 
viii. 33. 10 Indra is said to be Vrisha in the pardcat and 
also in the arudvat regions. Indra is again very often 
spoken of as going to distant regions to see whether Vyitra 
s duly killed. The same fact appears to be here expressed 
in a different form, 

VERSE 3. Sayana following the Armkramani, under- 
stands the verse as addressed by Indrani to Indra. Ludwig 
and Grassmann, on the other hand, take ifc to be addressed 
by Indra to Indrani: and this construction seems better 
than that of Sayana. It may, however, be here, once for all, 
remarked that though scholars thus differ in assigning 
verses to different deities, yet it does not, on the whole, 
materially alter the legend incorporated in the iiymn. 

* If Vrishakapayl is to be at all introduced in the dialogue, we 
may assign this verse to her. The phrases, " my friend " and "did 
not respect Indra, " would be more appropriate ia her mouth than 
in that of Indra or Indrami. 

Y( VWji" ;/,';,';" "'''"Jfe^i' 1 ''V, 


Says Indra: " What lias this Vrish&kapi, in the form of a 
yollow antelope, done to thee that you are so much angry 
with him ? Was it the rich possession (wealth) of the 
Aryans ? Indra, &c." 

The form, in which Vrishakapi is here said to have 
appeared, should be specially noted. Harita means yellow, 
and yellow animals (Haritah) are said to be yoked to the 
carriage of Aditya in Nighantu (]. 15). There the word is, 
however, understood to be the plural of Harit, by the com- 
mentators in conformity to Rig. i. 115. 3 and v. 45. 9, 
where the sun is said to have seven horses yoked to his 
carriage. But I think that the same idea may give rise to 
the conception that the sun is represented by a single 
yellow animal, and we may take the passage in the Nighantu 
as referring also to the verse under consideration. I have 
previously -alluded to the fact that the dog at the Chinvat 
bridge in the Parsi traditions is described as zaritem, that 
is, of the same colour as the antelope in the third verse. 
But the question of coloar cannot be taken as finally settled 
until we first definitely decide what animal is represented 
by Mriga.* 

VERSES 4 & 5. Sayana is literally correct, but again 
misses the spirit, or rather has missed it throughout the 
hymn. Indra was reproached in the second verse for his 
partiality or over-kindness to Vrisbakapi. But Indrani was 
not satisfied with it, and if Indra failed to punish the Kapi, 
she took the matter in her own hand. Says she : " Indra! 
Us you (thus) protect this (your) favourite Vrishakapi, let 
the dog, eager (to chase) a hog (vardha)> bite him at his 
ear. Tho Kapi spoilt my favourite things, f I shall, there- 

* See Dr. Rajendralal's Indo-Aryans, Vol. II., p. 303. 
f The word in the original is tashtani, \vhich literally means 
Wade, shaped, &c. Madhava Bliatfca understands it to 


fore, cut off his head, in order that an evil-doer may not 
enjoy happiness. Indra is in the upper (portion) of the 
universe." Here Indrani is herself prepared to punish 
VrisMkapi by setting the dog at his ear, and cutting oft' his 
head. I have in a previous chapter shown how the fig-are 
of Mriga's head is to be obtained in the sky. Taking the 
three stars in the belt of Orion as the top of the head, the 
dog is close by the right ear of Mriga and may properly 
be said to bite it. The word vardha also points out the 
place where we may expect to find the dog. In Rig. i. 
61. 7, vardha* is said to be killed by Vishnu beyond a 
mountain, which, in all probability, is the same story as that 
of Indra killing Vritra. A dog chasing vardha is therefore 
no other than Canis Major following the constellation of 
Orion, or the "antelope's head" representing Vritra. Say.-ma 
and Yaska, and even European scholars are silent as to who 
this dog is. The verses, in fact, may be said to have 
remained altogether unexplained hitherto, though the 
words themselves are simple enough and have caused no 

VERSE 6. This verse presents no difficulty. Thus 
satisfied, Indrani speaks of herself as the best of women, 
"best in every way. 

VEESES 7 & 8. Indra now tries to conciliate her. Sayana, 
following the Anukramani, supposes that the seventh verse 
is addressed by Vrishakapi and the eighth by Indra. The 
only reason I can find for such an interpretation is the 

oblations offered to lu'dranl. I translate it by things generally, 
"Whatever meantng we may adopt, it is quite evident that the Kapi's 
interfering with them has offended Indranl 

* In Rig. x. 99. 6, Indra is said to have killed Trishirshan, and 
with his aid Tritfc killed varaha. 


occurrence of the word ambd, which means cc mother/ 'and 
this cannot be supposed to be used by Indra, But though 
we avoid one difficulty in this way, we are launched into 
another, for the verse speaks of Indrani being pleasing 
" to me;" anil if Vrishakapi is the speaker (C me," cannot 
refer to him, as ludrani is his mother and, consequently, 
''me" has to be interpreted to mean (< my father/ 7 and this 
Sayjina has done. I prefer taking ambd as an affectionate 
and respectful mode of address, as in modern Sanskrit,, and 
the verse presents no difficulty. We can then take both 
the 7th and the 8th verse together and give them a natural 
interpretation. I translate thus " 0" auspicious lady! what 
you say is true you are pleasing to me . . 

But oh! hero-wife, with beautiful arms, pretty figure, 
profuse hair, and broad hips, -why should you be so angry 
with our Vrishakapi ? Indra is in the upper (part) of the 

V^KESK 9. Indrani replies, "This mischievous (Vrish- 
akapi ) considers me to be avird ( i.e., without a brave 
husband or son), while I am the wife of Indra, the mother 
of the brave, and the friend of Maruts. Indra, &c." 

V BUSES 10 & 11. Pischel and Geldner suppose that the 
first is addressed by Vrishakapi to Indrani, and the second 
by Vrishakapayi. Sayana understands them, to be addressed 
by Indra, Whichsoever construction we adopt, the mean- 
ing remains the same. Indrani is here told that she is 
highly respected everywhere ; she is the blessed of all 
women, and that her husband never suffers from old age. 
This is obviously intended to pacify her. 

VERSE 12. Indra says "0 Indrani ! I am not delighted 
without my friend Vrishakapi, of whom these favourite 


watery oblations reach the gods. Indra is in the upper 
( part ) of the universe." 

VERSE 13. This seems to be also addressed by Indra to 
Indrani, who is here called Vrishakapayi. This latter name 
has caused a difference of opinion, some considering 
Vrishakapayi to be the mother, some the wife of Vrishakapi.* 
I do not see how the wife of Vrishakapi, as such, could be 
introduced in the song, unless Vrishakapi is understood to 
be the name of Indra himself. Commentators, who take 
Vrishakapayi to mean the wife of Vrishakapi, accordingly 
adopt the latter view. Pisehel and Geldner think that tho 
verse is addressed by Vrishakapi to his wife Vrishakapayi. 
The verse means, u rich Vrishakapayi ! having a good son 
and a daughter-in-law, let Indra swallow the bulls, your 
favourite and delightful oblation. Indra, &c." There has 
been much speculation as to who could be the son and tho 
daughter-in-law of Vrishakapayi, But if Vrishakapayi be 
understood to mean the wife of Indra, it causes no such 
difficulty. The adjectives " having a good son," &c., are 
simply complimentary, corresponding to the statement of 
Indrani, that she was the <c mother of the brave" in verse 9. 
Indra accepting her statements, asks her to allow him 
to swallow the watery oblations said to come from Vmh- 
kapi in the last verse. The words priyam and liavis are 
.the same in both the verses ; and I think that both of them 
refer to the same oblations. 

, VERSE 14. Indra, satisfied with the prospect of getting 
the obMipns, describes his appetite 1 : t6 Twenty and fifteen 
oxen are being cooked for me ; I shall eat them and be fat. 

Both the sides of my belly will be filled up. Indra, &e^ 

* See Max Miiller's -Lectures on the Science of Language, 
Vol. 1L, &S8, * . 

// , 



The practice of sacrificing bulls to Indra seems to have been 
out of date even at the time of the Rigveda (cf. i. 164. 43, 
where it is said to be an old custom). But the old custom 
could not be entirely forgotten, and if real bulls were not 
offered to Indra, poets supposed that clouds or stars might 
answer the same purpose. The number 35 mentioned in the 
verse may thus refer to the Nakshatras (28), and planets (7). 
But this explanation is doubtf ul, and I cannot suggest a 
better one. 

VERSES 15, 16 & 17. The fifteenth and the sixteenth seem 
to be addressed to Indra by Indrani, and the seventeenth to 
Indra ni by Indra. In the fifteenth Indrarii, according to 
Sayana, asked Indra to sport with her just as a bull, with 
pointed horns, roars amongst a number of cows. The next 
two verses do not appear to be relevant to our purpose. 
We may therefore pass these over, and resume the thread 
of the story. Pischel and Geldner suppose that the 17th. 
and 18th verses are addressed by Vrishakap&yi. 

VEUSES 18 & 19. Indrani is now conciliated, and says 
that she has not killed Vrishakapi, but some one else. 
The verse thus means, " Indra ! let Vrishakapi get the 
slain animal an animal which was quite different from 
VrishAkapi's. Let him at once have a knife, a fire-place, a 
new vessel, and a cart-load of firewood (to cook the killed 
animal ). Indra/&c." Thus by the intercession of Indra, In- was moved, and at last undid or rather explained away 
her previous act of decapitation. Pischel and Geldner trans- 
late the verse very nearly as I have done. They, however, 
consider it to be addressed by Vrishakap&yi and translate 
parasvantam by f wild. 3 This does not explain what dead 
&airnal is here referred to. It is, I think, more natural to 
suppose that the dead animal here spoken pf is the same .a& , 



that described in verse 5, and one whose head Indram is 
fchere said to be ready to cut off. Indram now says that this 
dead animal should be given to Vrishakapi, especially as Indra 
has already got hie oblations of balls. I have already shown. 
that there were several legends about the " antelope's head." 
It seems that Indrdni, referring to some of them, assures 
Indra that it was not Vrishakapi in the form of the antelope 
which she killed, but some one else ( literally parasvan- 
ton=represecting another than Vrishakapi, as Say ana takes 
it). Thereon Indra, having thus saved Vrishakapi by his inter- 
cession, observes, " Thus do I go seeing and discriminating 
between a ddsa and an drya; I take my drink from those 
that prepare Soraa juice and cook the oblations, and thus 
behold or protect the intelligent} sacrificers. - In another 
word, Indra is glad that he has saved an Arya, and trium- 
phantly declares that he is always careful to distinguish 
between an Arya and a Dasa, the latter of whom he would 
punish and kill, e. g., Vritra, who is said to be a Dasa. 
Vrishakapi being thus saved Indra, in the following verses, 
"bids him a farewell, wishing for a safe journey and speedy 
return. These verses are very important for our present 
purpose, and I shall therefore examine them singly. 

20. In this verse Indra asks Vrishakapi to go 
to Ms house (astam) and then return afterwards to the 
house ^nhas) of Indra. But the question is where is 
Vrishikapi's house and where is that of Indra ? The words in 
the original are dhanva, krirttatra and nediyas. Vrishlkapi 
' is asked logo iodhanva, which is also krtotatra. Sayana 
takes .d&owwto mean, a desert and krintatra in the, sense 
that "the treejs therein are cut off.'* .But this meaning 
does not qwte suit the context. What is meant by- saying 
that VriMkapi, .wia? is admittedly the sun in- a 


form, should go to a forest? Where is that forest, and 
what does it imply ? DJianva is a word that occurs several 
times in the Rigveda. In Rig. i. 35. 3 it is saiu to consist 
of three yojanas and is contrasted with the earth. Sayanai 
there understands it to mean " sky or heavens ; " and I see 
no reason why we should not interpret the word in the 
same way in this verse, Dhanva therefore means "gkj' J or 
" heavens." Bat is it the vault above with three stages ? 
No, the poet qualifies the idea by tyintatra, meaning "cut 
off." It is thus evidently the portion of the heavens 
which is cut off. In other words, the idea here denoted 
is the same as that expressed by the phrase avarodhanam 
divah ee where heavens are closed/' or " where the view is 
obstructed," in Rig- ix. 113. 8. Dhanva, which is kritantra,* 
thus denotes the innermost part of the celestial sphere, 
the southern hemisphere or the Pitriyana. The poet knows 
that the vault of the heavens above him has three halts or 
stages which Vishnu is said to have used as his three steps 
(Rig. i. 22. 17). But of the nether world the poet has 
no definite knowledge, and he therefore cannot specify the 
yojanas or the stages it contains. Thus he simply says that 
there are some yojanas therein. The first part of the verse 
may now be translated thus : " Vrishakapi ! go to the 
house (in) the celestial sphere which is cut off and which con- 
tains some yojanas or stages/' In short, Indra means that 
Vrishakapi should now descend into the southern hemisphere. 

The latter part o the verse literally means " and come 
to our house from nediyas." Now nediyas. is again a 

* The only other place where "krintatra is used iu the Rigveda is 
v. 27. 13, which Yaska and Sayaaa both interpret to mean that 
" waters come up from krintatra, i.e., a cloud." But it may be 
as well asked if Itrfatolrdt cannot here mean " from below." 


word which neither Yaska nor Sayana seem to have properly 

understood. Panini (v. 3. 63) tells us that nediyas is the 

comparative of antika. Now nediyas catmot possibly be 

derived from antika by any change in the form of the latter 

word. Panini therefore considers neda to be a substitute 

for antika, when the comparative form is to be derived. 

This is equivalent to saying that ( bet J is to be substituted 

for c good ' in deriving the comparative form of f good * in 

English. I need not say how far such an explanation 

would be regarded satisfactory. My own view is that 

nediyas had lost its positive form in the times of P&nini, or 

perhaps its positive form was never in use like that of 

' superior^ in English. But P&nini, who > as a grammarian, 

felt bound to account for all the forms, connected nediyas 

with ^antika, probably because the ordinary meaning of 

ned^yas in his time was the same as that of the comparative 

form of antika. But we cannot infer from this that nediyas 

might not have m^ant anything else in the days of Panini. 

Panini might have taken into account only the most 

ordinary sense of the word, and finding that a positive form 

was wanting connected it with the word which expressed 

the ordinary meaning in the positive form. The fact 

that Panini considers nediyas as the comparative of antika 

does not therefore preclude us from assuming", if we have | 

other grounds to do so; that nediyas originally meant some- 

thing else in addition to its present sense ; for Pfinini speaks 

of -the form and not of the meaning of nediyas. Having 

thus, shown that the authority of P&nini is not against me , i | 

I shall now give my meaning of nediyas. I think it means j 

lower, being akin to neath, beneath, nether* and corres* 

, * Bopp derives O.H.G. nidar from Sk. ni down, and dis- 
approves Grhnm's suggestion that it should be trace to a Gothic 




ponding words in other languages. The suggestion, I 
know, will be received by some with surprise and suspicion, 
and I must give my grounds for proposing a new meaning. 
There is no passage in the Rigveda where the use of 
nediyas might be considered as definitely deciding its 
meaning. In Rig. v. 52. 6, viii. 26. 10 and x. 101. 3, 
nedisJitlia or nediyas might be supposed to mean lowest or 
lower. But- the passages are not conclusive on this point, 
as the word there used might also be understood to mean 
' nearest/ ' nearer/ according to Panini. In the Brahmanas 
we 5 however, meet with more decisive passages. Thus in the 
Aitareya Brahrnana vi. 27 nediyas is contrasted with upari- 
slitd-t* Bohtlingk and Roth give a passage from the 
Kathaka recension of Yajurveda (28. 4), which says, " he 
ascends (drohati) to the heavens from HienethisJitha world." f 
Here the word ' ascend ' clearly shows that the nethshitha 
world must be understood to mean the c lowest world/ 
* world at the bottom. ' In the Tandya Brahmana (iii. 4, 2, 

verb ntihan, nath, nethvm, and divided as nid-ar, ar being a 
comparative termination. (Bopp, Com. Gr. Eng. Tr. 1860, Vol. I., 
p. 382). K. Brugmann compares Sk. ned'yas with Av. nazd-yah 
meaning * nearer,' and derives the words from nazd (ni down and 
sed to sit). Of. Sk. mda Lat. nldu, O. Ir. net, 0. H. U. nest=. 
a resting place (Comp. Gr. i., 51)1, ii. 4, 135). Both Bopp and 
Brugmarm do not propose any new meaning of nedtyas. But 
it is evident that whichsoever derivation we adopt the word is 
connected with ni down, and if we find passages in the Brahmanas 
where it is contrasted with upari-shtdt, we can, 1 think, safely 
understand ntdiyas to meau * lower* as suggested by its etymology; 
* nearer ' is a secondary meaning, ^ 

t Ait. Br. vi. 27. 

I Kathak;28. 4. ; 


3 3, 2) there occurs a passage where the directions for 
lowering the tone are given as follows : " Just as after 
creeping up to the top of a high tree (a man) gradually 
co.oes lower and lower so, &c."* The word for lowering in 
the text in nediyas sankramdt, and there is no possibility of 
mistaking its meaning. In the T&ndya Brahmanaii. 1. 3 the 
raising of the tone is described as ascending from top to 
top (agrdt agram) ; and nediyas sankrama must, therefore, 
mean a gradual lowering of the Yoice. In fact, nediyas 
sankrama represents the same idea as low-er-ing, that is, 
not taking a sudden leap down but descending from the 
highest point to the next lower, and so on. In all these 
places Sayana explains nediyas as meaning * nearer J accord- 
ing to Panini ; but in every case he has to strain the words 
to suit the context. It was not, however, Rayaiia's fault; 
for after nadiijas was once assigned to antika, all traces of its 
old meaning were naturally lost, and none dared to question 
Panmi's authority. Bat we now know that in other 
languages neatfi means low, and in several passages in the 
Brahmanas, we find nediyas contrasted with * upper * .or 
*top. ; This, in my opinion, is sufficient to prove that nediyas 
meant lower in the Vedic times. I have already shown 
that the authority of Panini is not against understanding 
the word in this way. All that he has laid down is that 
nediyas having no comparative form should be derived from 
antika without saying whether nediyas was or was not used 
in any other sense. I am therefore inclined to think that 
necliyas might have had more than one meaning even in 
JPanini's time, but he took the most ordinary meaning and 
derived the comparative form from antika. This in course 

t KB. Br. iii. 4. 2. 


of time served in its turn to restrict the denotation of the 
word only to one meaning, viz.* ' nearer.* 

I would therefore translate the verse thus, " Vrisha- 
kapi ! go to the house the celestial sphere which is cut off 
land which contains some (unknown) yofanas or stages. 
From your nether house come to our house. Indra is iti 
the upper (portion) of the universe." Nediyas is thus 
contrasted with uttara in the burden of the song. Both 
are comparative forms. Indra is in the uttara (upper) 
regions, while Vrishakapi is going to the nectiyas (lower) 
world ; and Indra expects or rather requests Vrishakapi to 
come back agaia to his (Indra's) house. That is the gist 
of the whole verse. The idea that the sun falls down from 
the autumnal equinox is an old one. In Ait. Br. iii. 18 and 
in Taifct. Br. i. 5. 12. 1 the ceremonies on the Vishuvan or 
the equinoctial day in a satra are described* and there w 
are told that "gods were afraid of the sun falling down 
from the sky and so supported him, " and being thus sup- 
ported he " became uttara to all." The Ait. Br* iii. 18 has 
thus the same word uitara that we have in this verse, and 
it is natural to suppose that both relate to the same subject. 
I have also quoted a passage from the Aitareya Brahmana 
where nedtyas is contrasted with uparishtdt. From these 
I infer that the verse, we are now considering, describes 
the descent of the son into the southern or the lower 
'hemisphere, and that Indra asks him to come back again to 
the house of gods, 'i.e., the northern hemisphere. I have 
already given in full my reasons for understanding nediyas 
in a different sense. But I may remark that, even accept- 
ing the common meaning of the word, the verse may still 
be interpreted in the way I have suggested. 

VERSE 21. Vrishakapi has gone down to the nether 


world. This verse now describes what Indra will do when 
he returns. Says Indra, <c "0 Vrishakapi ! you, the destroyer 
of sleep, who are going to the house, come back again, 
again by (your) way. We would perform the sacrifices. 
Indra, &c/' The verse thus distinctly refers to the recom- 
mencement of the sacrifices in the Devayana or the Uttara- 
yana as understood in old days. The word swvita is from 
the same root as vaitdnika and 'kdlpaydvaliai is from Jcrip, 
the root which gives us the word kalp a, in Italpasutras. 
Suvita Jcalpaydvaliai thus means * s we would perform 'the 
mitdnika ceremonies/' which, as described in the first 
verse, were stopped when the sun went down to the nether 
world. I may also here point out that the house in the 
i&tli&r world or,- as S&yaiia interprets it. the house of the 
Hei&y is called o$a literally f thrown/ while Indra's house 
is called griha. The sun goes down to the asta and returns- 
up to the griha of Indra. This verse, insofar as it speaks 
of the recommencement of sacrificial ceremonies., confirms 
the interpretation I have proposed for the preceding verse. 

YBRSB '22. This is the most important verse in the 
whole hymn. It " describes the circumstances under which 
Vrishikapi will return to Indra's house. Literally rendered 
It means, f O mighty Vrish&kapi !* when, you rising tro^. 

* Pisdbel and Geldner suppose that the Terse is addressed "bf % 
third person to TrisMkapi and ladra, probably because both, tliea^ 
names occur in the vocative case and the verb is in plural In that 
e$se the verse would mean, i When Indra and VrisMkapi 
both be in the Ixouse, where weuld the sintiing Mriga be, &c. 
This interpretation does not, however, make any change in die 
of the verse 'material for ear purpose. For whichsoever 
straciik>n we aiopl fk& ^n^a^D-a still rentairis yihy" JB the 
invisible when both Indra and Vrisbakapi at? tc 


wards (or rather northwards) would ome to (our) house/ , 

wherfe would that great sinner Mriga be ? Where he, who 
misleads people, would go ? Indra, &c." Now Yaska, in 
whose days all traces of Canis being once a star in the 
heavens were lost, could not understand what to make 
of the statement " where would that great sinner Mriga 
be ? " It means that Mriga would not be seen, would 
not be visible, when Vrishakapi goes t6 the house of Indra ; 
but "Yaska did not perceive what was intended by such a 
statement. He could not conceive that the constellation 
of Mrigashiras would be invisible, when the Sun in his 
upward march would be there at the beginning of the 
Devayana, that is,- when he comes to the house of Indra, 
and therefore he proposed to interpret Mriga in the sense 
of "the sun" (Nirukta 13. 3). Mrig-a/ says he, isderived from 
mrij to go, and means " going/' " one who goes and goes 
and never stops," in other wordsy " the sun." Now, says 
his commutator,- when a person goes into a house he can- 
not be seen by the outsiders. So Vrishakapi, when he goes 
to the house, cannot be seen by the people on the earth I 
I do not think that I need- point out the highly artificial and 
inconsistent character of this explanation The word Mxiga, 
so fa-r as I know, ite no 1 where used in the Eigveda in this 
sen&e.- Ag-ain,ifthe word Mriga in the third verse of this 
hymn- is to be understood as r meaning an antelope, is it not 
natural enough to suppose that the same Mriga, is referred to 
in this verse ?- Then,- again, how can the sun- be said to 
, become invisible to- the people when* he is in- the house 
of godsj Nor can he be invisible to Indra whose house he 
enters. What can, in such a case, be the propriety of the 
word: utdanchal or " rising upwards " ? If Mriga means the 
sun according to Yasfe% we shall have to suppose that the 



rising sun was mwisible, a clear contradiction in terms. I am 
sure Yaska here tried to explain away the difficulty in the 
way as he has done in the case of Vrika. But., in the 
present instance, the solution he has proposed is, on the face 
rf it* highly inconsistent, so much so that even S&yana does 
follow it. Sayana, however, has nothing else to propose, 
lie quietly leaves the word Mriga as it is and unexplain 1 - 
d in his commentary. In short, both Sayana and Yaska 
have found the verse too difficult to explain. The meaning 
I have proposed explains the verse in a natural and a 'simple 
manner, mud further corroborates the statement in th$ 
Rigged* previously referred to, viz., " Canis awakened 
at the end of tlie year." In tne Taittiiiya 
S. S. l f we are told that the Vedic priests^ 
HL f ^ observed the position of the son amongst 

in ihe morning, and, as the Naksliatras disappear 
(he sun rises, they determined the position by observ- 
ing Naksh&tra rose a little before tne snn. The 
tecords an observation to make which no 
is required. B tells tis that when Vrfghakapi 
to tlte of ladra Ms Mrig^ wad not visible any- 
indicating thai tbe sun rose irith Orion 
wc^l fe es^m^ally remarfeabM 
Wfe stttt most fe ftK;ia wlien he goes- td 
%MA^ life fewnien o tibe song- tells i& 
4* tk0> pinri of 

m wey e 



I*"*! . r 4fi ",' 

, " *'^y '/' 


(tNirukta 12; 12 ). If the meaning I" have proposed is 
correct, we have here a record of the position of the sun at 
th0 vernal equinox. I take pulvnghas in the original to 
mean *' great sinner ; " but it may be translated as Yaska 
proposes by " omnivorous ' ; or " voracious." But in either 
case I would take it as referring to the antelope's spoiling 
the things of Indrani. The point is that the sinning Mriga 
would not be with Vrishakapi when he again goes to the 
bouse of Indra, and Indrani would have no cause to com- 
plain of the presence of the odious Mriga at the time. 

VERSE 23. Sayana translates, " O arrow ! Ma-na's daugh^ 
ier, named Parshu, gave birth to twenty (sons) together. 
Let her whose belly was big be happy ! Indra is in the uppetf 
(portion ) of the universe." I cannot, however, under" 
stand what it means* Parshu, according to Sayana, is a Mrigi 
or a female antelope. Bat why address the arrow to give 
happiness to her ? Can it have any reference to the arrow 
With which Orion was killed ? Then who are these twenty 
^ons ? Are they the same as twenty mentioned in verse 14 ? 
In it likely that twenty alone are mentioned leaving the 
additional fifteen to be understood from the context ? The 
eoncluding verse undoubtedly appears to be benedictory. 
Buir I 'have not found a satisfactory solution of the above 
questions. Perhaps, -bhala meaning ' auspicious/ may be 
used for Vrishakapi, and Indra addressing him- pronounces 
benediction on the female that gave birth to the yellpw 
antelope and several dther stars that are supposed to be 
either killed or . swallowed by Indra in this hymn. But 
I cannot speak with certainty on tie point and must leave 
tHe Verse as it is. ' . 
; Jfaw let as see what are th^e leading features of the ; sfcor j 



that scholars differ in assigning the verses of the hymn to 
the different speakers, and here and there we meet with 
expressions and words which cannot be said to be yet satis- 
factorily explained. Some of the interpretations I have 
proposed may not again be acceptable to all. But these 
difficulties do not prevent us from determining the leading 
incidents in the legend, which may therefore be summarised 
sorctewhat as follows. Vrishakapi is a Mriga, and sacrifices 
are stopped where he revels. He is, however, a favourite of 
Indra, and consequently the latter, instead of punishing, 
follows him, Indrani, who has frerself been offended by 
the 'Kapi, now reproaches Indpa for his overfondness for 
the animal and threatens to ppnish. tfce beast by cutting off 
his head and letting loose a dog at his ear. Indra intercedes 
and Indian} assures bim that the punishipent has not been 
inflicted on bis fayorite beast, but on someone else. 
Vrishakapi is now going down to l}.is house and Indra, in. 
bidding farewell to Ids friend, asks him to come up again 
to his ( Indra's ) house, so that the sacrifices may be 
recommenqed ; and, strange to say, that when Vrish&kapi 
returns, in his upward march to the house of Indra, the* 
impertinent Myiga. is no longer to be seen J Vrishakapi 
Indra and Indraiii thus finally meet in the same house, 
without the offensive beast, and th@ hymn, therefor e con-* 
eludes ;v?ith a benedictory verse, 

There pan be little doubt that the kytpn gives a legend 
urrentin old Vedic days, But no explanation has yet 
been suggested, which accounts fpr all the incidents in the 
story or explains how it originated. VrisMkapi is a Mriga ; 
p,nd his appearanpe and disappearance mark the cessation 
and the recommencement of the sacrifices. The Indian tradi- 
tion identifies him ^ith the #un in on form or anoth^- and 


comparison with Greek Erikapaeos points to the same conclu- 
sion. Our Vrishakapi or Mriga must again be such as is 
liable to be conceived in the form of a head cut off from the 
body, and closely followed by a dog at its ear, unless we are 
prepared to treat the very specific threat of Indrani as mean- 
ingless except as a general threat* All these incidents are 
plainly and intelligibly explained by taking Vrishakapi to 
represent the sun at the autumnal equinox, when the Dog- 
star or Orion commenced the equinoctial year ; and, above 
all, we can now well understand why Vrishakapi' s house is 
said to be low in the south and how his Mriga disappears 
when he goes to the house of Indra a point which has been 
a hard knot for the commentation to solve. I, therefore, 
conclude that the hymn gives us not only a description of 
the constellation of Orion and Canis (verses 4 and 5), but 
clearly and expressly defines the position of the sun when 
be passed to the north of the equator in old times (verse 22) ; 
3,nd joined with the legend of the Ribhns we have here 
unnaistakeable and reliable internal evidence of the hymns 
of the Rigveda to ascertain the period when the traditions 
incorporated in these hymns were first framed and con- 
ceived. In the face of these facts it is impossible, to hold 
that the passages in the Taittiriya Sanhita and the Brah- 
manas do not record a real tradition about the older begin- 
ning of the year, 


Jtesulfcs of previous chapters Winter solstice in Phalguna and Magha 
Successive year-beginnings- in old times stated and explained The 

4 second traditional year- beginning in the Taittiriya Sanhit Winter 
solstice in Chaitra and vernal equinox in Punarvasu Vedie traditions 
corroborating the same The commencement of the sacrifice with 
Aditi, the presiding deity of Punarvasft The Abhijit day The 
asterismal Prajpati with Chitr for his head THE CONCLUSIONS 
Periods of ancient Vedic literature stated and described The Pre- 
Orion Period, 60004000 B. C, The Orion Period, 4000^500 B. 0. 
The Krittika Period, 25001400 B. C. -Pre-Bud-distic Period, 1400 
5DO-B. C. Not inconsistent with the results of Comparative Philo- 
logy or Mythology Bate of the precession of the equinoxes Correctly 
determined by the Hindus Continuous record of the different positions 
of the equinoxes in Sanskrit literature Traditions based upon the same 
Prajapati, Rohinl and Kudra Meaning of Kohin! The Krittik^s in 
the Taittiriya SanhitS, and Vedanga Jyotisha The equinox in Ashvint 
in later works Story of Vishvamitra Notices of the recession of thj0 
rainy season from BMdrapada to Jyeshtha Conclusions shewn to bie 
consistent with the traditions regarding the antiqu-ity of Zoroaster and 
the Vedas. ,. 

WE have thus traced back one of the traditions about 
"Hie old beginnings of the year, mentioned in the Taittirfyk 
SlslM, to the oldest of the Vedic works, and what is -stiH 
more important, sb own that the Vedic traditions are* in 
this respect completely corroborated by the oldest records 
and traditions of the other two sections of the Aryan race- 1 - 
the Parsis and the Greeks. The traditions of each natiam 
taken singly may not be conclusive, but when, putting all 
these together and interpreting one set iri the ligkt of 
another, we find that directly or indirectly all point to th6 
same conclusion, their cumulative effeet- camnot' -but ^ 



conclusive. Scholars have already discovered the similarity 
between the traditions of the three nations, "but without 
any clue to the period when all the Aryas lived together, 
it was impossible to reduce all these traditions into a 
harmonious whole. The traditions of Orion, and especially 
its position at the beginning of the equinoctial year, da, 
however, supply such a clue, and with its help the mystery 
about the oldest periods of Aryan civilization is consider- 
ably cleared up. Thus if Orion is now no longer a hunter 
of unknown parentage, we need not also indulge in uncer- 
tain speculations about the foamy weapon with which Indra 
killed his enemy, or how the four- eyed dogs came to be 
stationed at the Chinvat Bridge, or why the Ribhus are 
said to be awakened by a dog at the end of the year. 

Astronomically the matter is as simple as it could be. 
All our measurements of time are directly based upon the 
changes in the positions of heavenly bodies* But there is 
no measurement of time, at present determined, which is 
longer than the period during which the equinoxes complete 
their revolution in the ecliptic. It is, therefore, the best 
measurement of time for determining the periods of anti- 
quity, only if wa iiave reliable records about the position of 
heavenly bodies in early days. Fortunately, such records 
of the time, when the Hellenic, the Iranian and the Indian 
lived together, have been preserved for us in, the 
, and. with the help of the Greek and the Parsi 
traditions we can now decipher these records inscribed on 
the specially cultivated memory of the Indian Aryans.' 
Commencing with the passages" in the Taittiriya Sanhita 3 
and the Brahmanas, which declare that the Phalgunl fall-' 
moons was once the new-year's night, we found that Mrig-a- 
&kiras was -designated by a aame which, if . rightly inlet-* 1 


preted, stowed that the vernal equinox coincided with, that 
asterism in old times. This was, so to speak, a sort of 
corroborative evidence of the truth of the statement in the 
Taittiriya Sanhita. A reference to the figure will show at 
a glance that if the sun be at the winter solstice on the 
FJialgunf fall-moon day, the moon to be full must be 
diametrically opposite to the sun and also near Phalguni. 
ITttara Phalgunt will thus be at the summer solstice and the 
vernal equinox will coincide with Mrlgashiras. With the sol- 
stice in Magha, the equinox will be in the Krittikas ; while 
when the Uttar&yana begins in Pauba the equinox is in 
Ashvinl, Ashviaf and Pausha, Krittikas and Magha, and 
MrigasMras and PMlguna are thus the correlative pairs of 
successive yeai^begmnings depending entirely upon the pre- 
cession of the equinoxes; and the facts, statements, texts and 
legends discussed in the previous chapters supply us with 
reliable evidence, direct and indirect, of the existence of all 
these year-beginnings in the various periods of Aryan cif ili- 
zation. It has been further shown that not only the traditions 
but also the primitive calendar of the Parsis bears oat the 
conclusions we have deduced from the Vedic woi*ks. 

We have eo far considered only one of the traditional 
year-beginnings recorded in the Taittiriya Sanhita, the 
Plialgun! full-moon. But it may be asked how we interpret 
the other mentioned along with it r and almost in the aurae 
worAs. Analogy at once suggests that we should interpret 
it in the same way as we have interpreted the first. With 
the Phalguni full-moon, at tlie winter solstice the vernal 
equinox was in Mrigashiras j, 80 with the GbifaA 
at the solstice tlte vernal equinox would be in 
Let us, therefore, see if we have any evidence IB Ob& . 
literature in support of such, an interpretal^>m* Ifc fee 


The figure is drawn on the supposition 
that the earth < E ) b i n . the cento, that the sna B10 ve 3 m ite 
Ecliptic, and' the precession of the eqxunoxes is caused by lie 
motion of the Ecliptic With a given Nakshatra at the ve ,rn a , 
equinox, we can hee at once find what Nakshatras wold Ipe at 
the other cardinal points and hence also the month ai the 
winter solstice, ;/ 



observed that we are here entering upon tie remotest period 
of antiquity, when the year was probably first determined 
with some approach to accuracy -j aad even in the Vedas 
there is hardly a-aythmg beyond vague traditions about this 
period, while tke Greeks and the Parsis have not, it 
appears, preserved eTCii these, 

There is no express passage which states that Punarvasft 
was ever the first of the Nakshatras, nor have we in this 
case a. synonym like Agrahdyana, or Orion, wherein we 
might discover similar traditions. There are, however, 
some indications about the oldest position of Ptmarvasii 
preserved in the sacrificial literature. The presiding deity 
<of Panarvasi is Aditi, and we are told in the Aitareya 
Brahmaaa i. 7, and the Taittiriya SanliiAvi. 1. 5. 1, that 
Aditi has been blessed with a boon that all sacrifices must 
oommemce and end with her. The story begins with the 
statement that the Sacrifice (the mysterial sacrificial per- 
sonage) went away from the gods. The gods were then 
unable to perform any further ceremonies, and did not know 
where it ( the sacrifice ) had gone to:; and it was Aditi tha-t 
helped them, in this state, to find out the proper commence- 
ment of the sacrifice.* This clearly means, if it can 
mean anything, that before this time sacrifices were per- 
formed at random, but it was at this time resolved -and fixed 
to commen'ce them from Aditi. Aditi was thus the oldest 
and the first commencement of the sacrifice or the year 
In the V&jasaaeyi Sanhita 4. 19 Aditi is said to be ubhaya. 
shlrshm, " double -headed/' and the commentators 

* Ait. Br. i. 7. A similar traditioa about Orion is narrated in 
Greek mythology. It is stated that having lost his sight he follow- 
ed a guide to the east in search of the sun and there, by exposing 
iais face to the rising sian, his sight was restored. 


; interpret it to mean that the two termini of the sacrifices, 

which began and ended with Aditi, are the two heads here 
| alluded to. These traditions are further corroborated by 

j the sacrificial ceremonies. According to the sacrificial 

! terminology the 4th day before Vishuvan or the central 

j day of the yearly satra is called the Abhijit day. " In the 

j sixth month/' observes Lr. Haug,* "there are three Abhi- 

\ plava, shalahas ( six-days' periods ) and one Prishthya 

| shalaha." This makes up the first 24 days of the sixth 

| fnonth. The following days are thus enumerated : "the 

I Abhijit day, the three svarasdman days and the Vishuvan, 

j or the central day which stands quite apart." Thus if we 

j exclude the Vishuvan day, as standing apart by itself, this 

5 gives us four days* and with the two days Atirdtra and 

1 Okafnnimha which are taken np by the initial ceremonies 

t^ pi the mtra, we make up the shalaha wanted to complete the 

! six months. The Abhijit day thus falls on the fourth day 

before the Vishflvan. Now if Abhijit day be supposed 
i\Gs "N* ^ e nam ^ after the Nakshatra of that name ( i. e., when 
^ |1the sun is in Abhijit) then the Vishuvan or the autumnal 
equinox must fall four days or as the sun travels over 
about 1 of the ecliptic each day, 4 af ter the asterisca 
gi Abbijit ; and it can be shewn by astronomical calculation 
ihafc, with Aditi or Punarvasfi at the vernal equinox to 
commence the sacrifice, we get nearly the same result* In 
the Sfiiya SiddMn.fca ( viii 3 ; table ) the longitude of 
Ponarvagi Is mid to be 93, while that of Abhijit is 268 
4fF, that is, la other words, Abhijit would be about 6 behind 
the asteranal equinox or Vishuvan, if we suppose the vernal 
equinox to exactly coincide with Punarvasfi. With the 

~~ ' "" ^~ --- __ : ; , ___ . - , . I 

* See Dr.. Stag's translation of the Aitareya Brahmana iv, 1^ 



vernal equinox in Panarvasu there is again no oilier Nak- 
shatra nearer to or at the autumnal equinox to mark the 
Vishuvan day. We can, therefore, now understand why 
Abhijit, which is so far away from the ecliptic, should have 
been included in the old list of the Nakshatras, It marked 
the approach of the Vishuvan in the primitive sacrificial 
calendar, but when it ceased to be used for that purpose 
owing to the falling back of seasons, it was naturally drop- 
ped from the list of the Nakshatras, as it was far away from 
the Zodiac. If Bentley's suggestion about Mula and 
JyeshtM be correct, this must have been done at the time 
when the vernal equinox was in Orion. But be that as 
it may, it will, I think x be clear from the above that the 
position of the Abhijit day in the sacrificial literature fully 
supports the tradition about Aditi, the presiding deity of 
Punarvasu, having discovered the commencement of the 
sacrifice. Aditi at this time must have also separated the 
Peyay&na from the Pitriyana and thus may have been 
appropriately called the mother of the Devas ( Rig. x. 72, 
5 ). * It was from her that the Adifcyas were born (Rig. x, 
72. 8; Shat. Br. iii. 1. 3. 2.), or the sun commenced his 
yearly course. 

the only other tradition I could find in the "Vedic litera- 
ture, about this position of Aditi is the story of the 

* Aditi is here said to be the daughter of Daksha, also cf. Rig. 
viL 66, 2. In Purauic traditions the 27 Nakshatras are said to be 
the daughters of Dakslia who gave them to the moon. If we com- 
Mne these two traditions Aditi would |>e at the head of all thfe 
Nakshatras, ia the same way as Mrigashiras or theKrittikas beaded 
the list m later times. There ate again many legends in the Pura- 
nas, stating that everything w&s horn from Aditi*. "We can accouxit 
for all these facts if we place -Adttr -at- the vernal equinox, when the 
calendar was first fixed f$ the jsamicial purposes. 



asterismal Praj&pati given in the Taittiriya Bralimana (i. 5 r 
2. 2).* The asterism of Chitra is here said to be the head 
of this Prajapati, Svati the heart, Ha-sta the hand, "VisMkha 
the thighs, and A nftrMhar the foot. .Many sonjeetures are- 
made about the meaning of this figure, but none of them 
satisfactorily explains* why Prajapati>, who is said to be> 
the god of time or the lustrum of years in the Vedanga 
Jyotisha, &hon*ld bave been represented in this way, I 
propose that we should interpret it after the manner of a. 
similar representation of Brahman by Badar&yana, t where- 
in the different signs of the Zodiac are said to be similarly 
related to the different parts of the body of Brahman r the- 
Creator. Prof. Max Miftller has thiit-s translated the- 
verse : " The ram is- the head, the face- of the Creator is the- 
bull,- the breast would be the man-pair,, the heart the crab, 
the lion the stomach, the maid the hip*, the balance-bearer 
the belly, the- eighth (scorpion) the membruna, tfoe archer his- 
pair of thighs r tlae Makara hi& pair of knees, the- pot his 
pair of legs, the fish his two feet.-"} Thias i Mesha was 
Brahman's head when the Eashis were introduced, 
Chitra could well be said- to be the head of Prajslpati when 
the Chitra full-moon commenced the year. But though w& 
ean thus satisfactorily account fo the- fact why Chiirl 
should have been called the head of Prajapati, yet we can- 


J India; what it caa teach us ? pp. 3-22, 323, 


not give an equally satisfactory reason in the case of one of 
the Nakshatras in this representation, unless we place three 
intercalary months in five years. It . is, however, very 
difficult to determine how the intercalary months were 
inserted, if at all, at this remote period, and the question 
must therefore, to a certain extent, remain unsolved for the 
.present. The analogy of the pictorial representation of the 
twelve signs of the Zodiac in later days, is, however, a 
strong ground to hold that the asterismal Prajapati may 
have been .similarly conceived when the primitive year was 
first determined on the Nakshatra system. There is, so far 
as I know, no more evidence about this primitive calendar 
in the Vedic works, than what has been given above. 
But the traces of such period which we can discover in the 
sacrificial literature and especially the express mention in 
Taittiriya Sanhita that the Chitra full-moon once commenced 
the year are, in my opinion, sufficient to prove the exist- 
ence of such a calendar in the primitive days* We 
cannot otherwise account why the first and last offerings in 
every sacrifice should be made to Aditi and why Abhijit- 
clay should precede the YisMvan by four days. Compared 
to the evidences of the Orion period, these are slender mate- 
rials for the construction of the primitive Vedic calendar, but 
they are decidedly superior to the materials on which 
Dr. Geiger has determined the primitive calendar of the 

It appears to me therefore that the oldest Vedic calendar 
like the oldest hymn, was sacrificial ; and that the sacrifice 
or the year commenced with Aditi at the vernal equinox in 
or near Punarvasu. The phases of the moon, the seasons 
and the ayanas further guided the ancient Aryas in 
measuring time for sacrificial purposes. The asterism of 
Abhijit marked the approach of Vishuvun or the central 


day, while Punarvasu, which soon after came to be called 
Yamakau, perhaps Yama and Yami, indicated the begin- 
ning of the year. Sometime after this and before the 
vernal equinox had receded to Orion, the Innar months and 
tiihis or days appear to have come in use ; and, in fact, the 
whole calendar seems to have been rearranged, the year 
being made to commence from the winter solstice in the 
Chitra full-moon. But this did not alter the sacrificial 
system, which, so far as the procedure is concerned, still 
Continues to be what it was in the oldest days. For all 
civil purposes the new calendar was, however, at once 
adopted and the two systems have, continued to exist 
side by side up to the present day> though in a consider- 
ably modified form, as described before in the second 

The oldest period in the Aryan civilization may there- 
fore be called the Aditi or the pre-Orion period, and we 
may roughly assign 6000-4000 B. C. as its limits. It was 
a period when the, finished hymns do not seem to have been 
known and half -prose and half-poetical Nivids or sacrificial 
formulae "giving the principal names, epithets, and feat$ 
of the deity invoked " were probably in use. The Greeks 
and the Parsis have retained no traditions of this period, 
for the simple reason that they carried with them only 
thQ calendar which was in force .when they left the common 
home, while the Indian Aryas have preserved all th 
traditions ....with a super-religious fidelity and scrupulous- 
sess. It is thus that I explain why. the oldest Grreek an;<d 
Parsi traditions do not go beyond Orion. 
,.. We next come to the. Orion period which, roughly speafcr 
irrg extended from 40,00 B. C* to 2500 B. 0., from the 
"when the vernal equinox was in the asterism 
time when it receded to -the astqrmnx Jof th. 


This is the most important period in the history of the 
Aryan civilization. A good many suktas in the Rigveda 
fag., that of Vrishakapi, which contains a record of the 
beginning of the year where the legend was first conceived) 
were sung at this time, and several legends were either 
formed anew or developed from the older ones. The Greeks 
and the Parsis appear to have left the common home daring 
the latter part of this period as they have retained most 
of these legends, and even the attributes of the constellation 
of Mrigashiras, otherwise called Agmyana, Orion or 
the Pauryeni. We can now easily understand why 
no confirmatory evidence about the Krittika- period is found 
either in the Eigveda or in the Greek and Parsi legends and 
traditions. This w,is pre-eminently the period of the hymns. 

The* third or the Krittika-period commences with, the 
vernal equinox in the asterism of the Krittikas and extends 
up to the peiiod recorded in the Yedanga Jyotisha, that is, 
from 2500 B. 0. to 1400 B. C. It was the period of the 
Taittiriya Sanhita and several of the Brahmanas, The 
Jiymns of the Rigveda had already become antique and 
unintelligible by this time and the Brahmavadins indulged 
in speculations, often too free, about the real meaning of 
these hymns and legends, attributing the use of the foamy 
weapon used by Indra to a compact between him and Na 7 
rmichi. It was at this time that the Sanhitas were probably 
compiled into systematic books and attempts made to ascei^ 
,tain the meanings of the oldest hymns and formulae. It wag 
also during this period that the Indians appear to tavo 
.come in contact wifch t]ie Chinese, and t^e latter borrowed 
the Hindu, Naksaatra system. I do ,not mean to , say. 
Hindus might not have improved, their ay stena ;by 
interchange of ideas as th,ey did wlien they camQ, t JtoQ w_o 


Greek astronomy. But the system was decidedly of Hindu 
origin and of purely Hindu origin being handed down from 
tke remotest or the pre-Orion period in the Vedic literature. 
M. Biot was unable to assign any reason why the Chinese 
should have taken a leap from the shoulder to the belt of 
Orion to choose their fourth $ieu. But with the older 
Hindu traditions the question admits of an easy explanation, 
as the belt was therein the real Mrigashiras or rather the 
top of Mriga's head. 

The fourth and the last period of the old Sanskrit litera- 
ture extends from 1400 B. G. to 500 B, 0. or to the birth 
and rise of Buddhism. It was the period of the Sutras and 
philosophical systems. It may be called the real pre-Bud- 
dhistic period, But as this has been sufficiently discussed 
by other writers I need not go into its further details. 

I do not mean to lay down hard-and-fast limits o each of 
these periods of antiquity, nor do I intend to say anything 
about the period which must have elapsed before the Vedic 
Aryas were able to fix their primitive calendar in the Aditi 
period. The beginning of the Aryan civilization must un- 
doubtedly be placed a long time before the people were able 
to conceive and determine the calendar. But I do not wish, 
to enter here into these speculations. I take my stand only 
upon what we find recorded in the Vedic works, and hence 
all that I mean is that if the astronomical allusions, references, 
facts, and legends in the Vedic works can have any meaning*, 
we cannot materially shorten the periods I have here in- 
dicated. We may not rely on vague traditional beliefs 
amongst one nation alone, but when we find that the tradi- 
tions of India, Greece, and Iran, agree in their important 
features, and can be explained satisfactorily only by placing 
the vernal equinox in Orion, and when we have an express 


authority for doing so in the Rigveda, I do not think that 
we can reasonably refuse to accept the conclusions deduced 
therefrom. It is true that we have determined the oldest 
Vedic periods from the traditions we find recorded in the 
Rigveda, and, strictly speaking, it is the period of the tradi- 
tions and not of the hymns into which they have been in- 
corporated. But this does not, in my opinion, materially 
affect the conclusions we have arrived at above regarding 
the ancient periods of the Vedic literature. I do not mean 
to deny that .the hymns. may not have been sung some time 
after these traditions and legends were originally conceived,, 
or that after they were first sung the hymns might not 
have been somewhat modified in form in passing from 
mouth, to mouth before they became settled in the form in 
which we now possess them. But though, so much, may 
be legitimately conceded, I think that it is impossible to 
hold that the hymns were composed thousands of years 
after the stories narrated in them, were first conceived* 
For, as a matter of fact, we find that the Rigveda hymns 
had already become antiquated and unintelligible in the 
days of the Taittiriya Sanhita and the Brahmanas. The 
,Taittir!ya SanhitS, places the vernal equinox in the Krittikas, 
and I have shown that we must fix its date at about 2500 / 
J3, 0. If the hymns of the Rigveda Sanhita were unintelligi- 
ble at this time, they must have been sung several centuries 
beforeiti The comparison of the Taittiriya with the Eigveda 
Sanhita further shows that while the first mentions three 
year-beginnings one current and two old the second 
.qp]y mentions one. Again, the Rigveda Sanhita contains 
up reference to the Krittikas as the mouth of the Nakshatras, 
I therefore conclude that the legends in question must have 
been incorporated into the hymns of the Rigveda* when 

: THE 

-were .still intelligible, that is, in the Orion period- 
it is of course impossible to determine the dates of indivi- 
dual hymns. That all of them were not sung at one time 
is quite evident from their style. Some of the hymns dis- 
tinctly speak of older hyinns or bards, while in Rig* x, 
90 '9 the hymns are said to proceed directly from ihejpurdsha 
or the sacrificial personage. All that we can therefore legi- 
timately say is that the hymns, which contain older traditions 
iand legends, e. g^ of the Ribhus and VrisMkapi, must have 
"been composed in the Orion period. Some of the hymns 
-may even be still older and some later, but generally speak* 
ing we may suppose that 4000 and 2500 B. C. are the limits 
of this period. This may require us to assume the existence 
bf some Vedio verses at a time when the Hindus, the 
'Greeks and the Parsis lived together. Some scholars may 
'hesitate to accept such a conclusion. But so far as I know 
the conclusion is not inconsistent with the results o com- 
parative Philology or Mythology, Prof. Max Miiller in 
Ms Biographies of Words ( pp.* 188-198 ) gives a Hat of 
'-about sixty mythological names which may bo shewn to b 
'common to Greek and Sanskrit.* If so many mytho- 
logical names can be shewn to be phonetically identical!, 
it is* impossible tD suppose that no songs, celebrating tbe 
"deeds of these deities, existed in the Indo-Germanic period, 
'Westphal has already proved the existence of poetry in th 
^Indo-Germanic period, and Dr. Kuhn has endeavoured tt 
! trace whole formulae back to the beginning of. Indo-Europeaii 
.* For instance Eidhu is cotnpared to Greek Orpheus, Saram& 
*to Gk. Menes, Vritfa to Gk. Ortlwos, Bdmliantw to &k.j)eopjken^ 
te$> I have already referred to his suggestion regarding the 
{ comparison of Friskmkcepi with &k. Jlrikapaeos* If all these 
^deities existed k the Indo-^Ger^anie period^ wij not 


poetry. Verbal coincidences such as, St. pada^Av. padha,, 
&k, pous, all meaning a metrical foot, again point to the 
same conclusion.* The results of comparative Philology, 
are, therefore, not only not inconsistent with, but, on the. 
contrary, corroborate the conclusions we have independently 
deduced from the astronomical references and allusions ; 
recorded in the old Vedic literature. But I would not make 
nay case rest on such grounds. It must be remembered, 
that we have not been speculating . in any way about the 
eldest Vedic periods. Our conclusions have been based on 
express statements and texts in the Vedic literature, and un- 
less the texts themselves are questioned or other more reason- 
able interpretations suggested, we shall not be justified 
in disregarding these results, simply because they do not 
support certain literary hypotheses, guesses, or conjectures, 
as, for instance, those that have been previously refer red^ 
to in the first .chapter. The results of the literary method 
may be moderate. But moderation is a virtue only when 
we have to make guesses about the periods of antiquity 
from uncertain data. Where however we have definite 
'texts, and traditions to rely upon, nothing biat prejudice 
pan deter us from drawing legitimate conclusions from, 
them on the ground that they take us too far back. The 
astronomical method, I admit, is vague, insofar as it does 
not enable us to determine the exact date of all the Vedic 
'hymns x>r ^rorks, but it is certainly superior to the linguistic 
method inasmuch as it supplies us with certain definite and 
pn disputed t f acts, for instance,, the position of the equinoxes 
which can safely be made the nuclei of the different periods 

* See Dr. Schrafc's Pre-historic Antiquities of Arjan Peoples,, 
I., CKap. IL, pp. 27, 28. 


of antiquity. "When the centres of each, period are tlius un~ 
disputably fixed and determined, we can then use the 
literary or the linguistic method to supplement these results- 
by determining the duration of each period. There would 
then be no real opposition between the two methods. The 
one would determine the specific points of time, while the 
other would give us the range of the different periods. In . 
other words, the first would supply the piers and the second 
the arches of the bridge, which we mean to construct across 
the period of antiquity, and which must therefore be com- 
pleted with the assistance of both. 

It may, however, be urged that if the beginning of the 
year was twice altered owing to the precession of the 
equinoxes, how is it that we do not find the traces of the 
intermediate stages or of the changes in the seasons in the 
old Vedic works ? How, it m^y be further asked, did the 
Indian Aryas not discover the precession of the equinoxes 
in the early Vedic times ? But it is not at all difficult to 
answer these questions. We might as well ask how no one 
before Bhaskaracharya or Newton ever thought of the 
attraction of the earth, though since the very beginning of 
the human race every one observed heavy objects falling 
down to the surface of the earth. The reason is plain 
enough. Celestial and natural phenomena cannot be fathom- 
ed or understood without a steady and close observation for 
centuries, and, above all, until all the auxiliary, or rather 
the whole group of sciences are proportionally developed. 
If we bear this circumstance in mind, we can, I am sure, 
discover sufficient traces of the intermediate changes in the 
Vedic works. Thus we find that of all the ancient 
nations the Hindus alone had well nigh accurately 
determined the rate of the motion of the precession, of the 


equinoxes.. Hipparchus considered it to "be not less than 
36", while the actual motion at present is 50*' 25 per year* 
Ptolemy adopted, as observed by Prof. Whitney, the 
minimum of 36" determined by Hipparchus; and it is 
evident that the Hindu astronomers who fixed the rate at 
54* per year could not have borrowed it from the Greeks. 
Prof. Whitney is at a loss to understand how the Hindus 
succeeded in arriving at a determination of the rate of 
motion, so much more accurate than was made by the great 
Greek astronomer, and he observes that it might be a 
4t lucky hit on their part/' * But why should they try to 
hit, even luckily, when they could have easily borrowed it 
from the Greeks ? I am therefore disposed to think that it 
was independently, and almost correctly, discovered by the 
Hindus long before other nations could do so, though we 
cannot exactly fix the period when it was done; and that 
there were sufficient materials for the purpose in the old 
literature of India. 

- Let us next see what traditions about the intermediate 
stages have been preserved. First of all I refer to the 
tradition of Eudra killing Prajapati, the god of time, for 
receding towards his daughter Rohini. The Aitareya 
Brahmana (iii. 33) describes this conduct of Prajapati as 
akrita or unprecedented and such as deserved to be 
severely noticed by the gods. Can we not herein discover 
the fact that the sun was gradually receding towards 
Kohini, by the precession of the equinoxes ? The ancient 
priests, who observed the fact as they watched the Nak- 
shatras at the commencement o the year, could not account 
for the change, and they rightly and honestly believed 
that it was a great calamity that the sun or Prajapati 
* See Whitney'* notes to the Siirya Siddhaiita^ iii. 13* p. 

214 ,TBE ORIONi 

should 'thtis follow an unprecedented course.. I Have 
previously referred to a verse from Garga,* which says that 
if the Otta'r&yana commenced otherwise than - from the 
;asterism of Dhanishtha, it foretold a great danger ; and wa 
may suppose that the Vedic Aryas similarly believed that 
if the sun ceased to commence the year from Orion, it was 
an unprecedented calamity. Prajapati, however, was punish- 
ed for .his unusual conduct, and there the matter ended foE 
the time being. I may also refer here to the ancient mode 
,of deriving the word Eohini. The Arabs called it A1-* 
JDabaran or/ 6 the follower*' evidently because it came next 
after the- Krittikas.f But the Hindus called it Eohini, 
"the ascended," inasmuch as they noticed that the sun 
gradually ran towards it in oldest days. It has been sug* 
gested that we should explain the legend of Prajapati by 
(reference to the daily rising of Echini, Mrigashiras, and 
Eudra in succession,. But this explanation hardly accounts 
for the fact why Prajapati was considered as literally run-* 
aiing after Eohini in an unprecedented way. Surely we 
Cannot suppose that the Vedic priests were ignorant of th.q 
fixed position of these constellations, and if so, we cannol^ 
jaccount for the fact why they considered Praj&pati as 
after 1 and thinking of living together with Bohinl 
they had, noticed the actual recession of the sun to^ 
wards Rohini owing to the precession of the equinoxes* 
ffhe tradition of Prajapati and Eudra, is thus comparatively 
speaking a later tradition^ though it seems to have been 
-completely formed before the separation, of the Greeks and 
efehe Parsis from the Indian Aryas, 

* See supra, Chapter II, p. 1% 



r fiut the question which was dropped" at this time after 
'punishing Prajapati, was again taken up when the equino 
tad receded to the Krittikas. The seasons had fallen back 
fcy one fall month and the priests altered the year-begin- 
king from the Phalguni to the Magha full-moon, while the 
list of the isfakshatras was made to commence from the 
Krittikas," instead of from Agrahayana. There is nothing 
surprising in the fact that the change should have been 
quietly introduced when we see that Varahamihira did the 
same in the fifth" century after Christ when the'Ashvini- 
!system was introduced.* The calendar was mainly used for 
the sacrificial purposes, and when the priests actually observ- 
ed that the sun was in the Krittikas, and not in Mriga* 
shiras, when day and night were equal, they altered the 
commencement of the year to the Krittikas, especially a& 
it was more convenient to d.o so at this time when the cycle 
>f seasons had receded by one full month. The priests 
sknew that the year commenced a month earlier in older days^ 
but like Varahamihira they must have appealed more to the 
actual facts., as they saw them, and introduced the change 
'without attempting to discover its real cause. } 

The Vedanga Jyotisha introduces the third change, whetii 
the season3 had further fallen back, not by a month, but 
by a fortnight. It was probably during this interval that 
the beginning of the month was altered from the full-moon 
to the new-moon, and when this beginning of the montli 
was so altered, advantage was taken of the receding 
of the seasons by a fortnight, to commence the year with 
the new-moon in Dhamsshtha as the Vedanga Jyotisha has 
done* . , . -^ 

* See supra, Chap. IIL, p. 36. 


., From this the next recorded step is to Ashvinl There 
is, however, an interesting story related in the Mahabharata 
which evidently refers to an abortive attempt to reform 
.the calendar when the seasons had again fallen back by a 
fortnight. In the 71st chapter of JLdiparva we are told 
that Visvamitra attempted to create a new world,* and 
make the Nakshatras commence with Shravana, instead of 
Dhanishtha; and the same story is alluded to in the Ashva- 
medha Parva, chapter 44. The tradition can also be found 
in other Puranas where Visvamitra is represented as 
.endeavouring to create a new celestial sphere. It appears, 
however, that he did not succeed, and the Krittika-system, 
as modified by the Vedanga Jyotisha, continued to re* 
gulate the calendar until the list of the Nakshatras was 
quietly made to begin, as noticed in the third chapter, with 
Ashvini in later times. 

We have thus an almost continuous record of the year- 
beginnings from the oldest time down to the present in the 
literature o India, and in the face of this evidence it is use- 
less to indulge in uncertain speculations about the antiquity 
o the Vedas. I have already referred to the occurrence of 
the pifri-pahsha, in Bhadrapada as a relic of the time when 
the year commenced with the Phalguni-full-moon. Our 
Shravani ceremony appears to have been once performed 
in Bhadrapada (Manu iv. 95) ; and as it marked the begin- 
ning of the rains,, when the herbs appear anew (Ashvalayana 

* Haha. Adi. 71, 34. 

f ^ ?r^r I 1 ^^f ^ssrarefqr^r i 
r 1 1 

and again in the Ashr. 44. 2. 


{Jrihya Sutra iii. 5. 2), we can here trace the recession of 
the rainy season from Bhadrapada to Shravatta, and from 
Shr&vana to Ashadha ( Sankhyayana Brahmana i. 3 ) and 
finally from Ashadha to Jyeshtha, as at present, thus fully 
corroborating the recession of the beginning of the year or 
the winter solstice from Chaitra to Ph&lguna, from PMlguna 
to M&gha, and from Magha to Pans ha. The evidence of 
the recession of the seasons Is not, however, -as complete as 
that of different year-beginnings, inasmuch as there are 
various local causes besides the precession of the equinoxes 
"that affect the occurrence of the seasons. The seasons in 
the Central India and Central Asia caanot, for instance, be 
the same, and if the Aryas came into India from the North- 
-West, the very change of locality must have caused a 
corresponding change in the seasons. The evidence of the 
change of seasons cannot therefore be supposed to be- so 
reliable and conclusive as that of the successive changes in 
the beginning of the year above mentioned. 

, Lastly, there remains only one question to be considered* 
Jb the Vedic period here determined consistent with the 
.traditions and opinions entertained about it by the ancient 
and modern scholars ? I think it is. I have already 
referred to the remarks of Prof. Weber who, though he 
regards the Krittika evidence as vague and uncertain, yet 
on geographical and historical grounds arrives afc the con- 
clusion that the beginnings of the Indian literature may 
be traced back to the time when the Indian and the Iranian 
Aryas lived together; and this opinion is confirmed by the 
fact that there are Yashts in the Zend Avesta which may 
be considered as "reproductions" of the Vedic hymns. 
Dr Hang considers - that- this condition may be satisfied 
if we place tho-b^ginmng-of the Vedic literature, in 2400 


B. C;* but lie was not cognisant of tlie fact that the 
vernal equinox can be shown to have been in Mrigashiras 
at the time when the Parsis and the Indians lived together. 
In the light of this new evidence, there is therefore no rea- 
sonable objection for carrying the periods of the Vedic 
literature further back by over a thousand years OF to about 
4000 B. C. This period is further consistent with the fact 
that in 470 B. C. Xanthos of Lydia considered Zoroaster 
to have lived about 600 years before the Trojan War (about 
1800 B. C.); f for according to our calculation the Parsis 
jnust have separated from the Indian Aryas in the latter 
part of the Orion period, that is to say, between 50*00 to 
2500 B. C ; while, if we suppose that the separation occurred 
at a considerably later date^ a Greek writer in the fifth 
century before Christ would certainly have spoken of it as 
a recent event. Aristotle and Eudoxus liav gone still 
further and placed the era of Zoroaster as much as 6000 to 
5000 years before Plato. The number of years here giveu 
is evidently traditional, but we can at any rate infer from 
it this much that at the time of Aristotle (about 320 B. C,) 
Zoroaster was considered to have lived! at a very remote 
period of antiquity; and if the era of Zoroaster is to be 
considered so old/ a fortiori] the period of the Vedas must 
be older still. Then we have further to consider the fact 
that an epic poem was written in Greek in about 9'00 or 
1000 B, C. The language of this epic is so rmlike that of 
the Vedic hymns that we must suppose it to have been, 
composed long time after the Greeks left their ancient home 
and travelled westward. It is not, therefore? at all impro- 
bable 'that they separated after the formation of the legends 

* Dr. Haug's Intr. to Ait, Br., p. 48. 

t See Dr. Haug's Essays on Parsis, p. 298, 


of Orion and before the vernal* equinox was in the Kritti- 
kas, that is, between 3500 to 3000 B. C. Finally, we can 
easily understand ' how the- acutest and most learned of 
Indian theologians and scholars believed the Vedas- to have 
come down to them from an unknown period of antiquity. 
A revelation need not necessarily be anddi, or without a 
beginning. The history of the Bible and the Koran shows 
us that a revelation can be conceived to be made at a par- 
ticular period of time. If so, the mere fact that it is be- 
lieved to be revealed does not account for the opinion 
entertained by the Hindu theological writers that the Veda 
lias come down to them, from times beyond the memory of 
man. Some of these writers lived several centuries before 
Christ, and it is quite natural ta suppose that their opinions 
were formed from traditions current in their times. The 
periods of the Vedic antiquity we have determined render 
such an explanation highly probable* According to the 
Christian theology, the world was created only about 4000 
years before Christ ; or, in other words, the notions of 
antiquity entertained by these Christian writers could not 
probably go beyond 4000 B. C. and not being able to say 
anything about the period preceding it, they placed the 
beginning of the world at about 4000 B. C. The Indian 
theologians may be supposed to have acted somewhat ia the 
same manner. I have shewn that the most active of the 
Vedi.c period commenced at about 4000 B. C., and- there 
are grounds for carrying it back still further. The form of 
the hymns might have been more or less modified in later 
times ; but; the matter remained the same, and coming down 
from such a remote antiquity it could have been easily 
believed by Jaimini, Panini, and the Brahmavadin of old to 
have been in existence almost from the beginning of the 


worldj or rather tlie beginning of all known tilings. We 
can thus satisfactorily account for all the opinions and 
traditions current about the age of the Vedas amongst 
ancient and modern scholars in India and in Europe, if we 
place the Vedic period at about 4000 B. C., in strict accord*- 
ance with the astronomical references and facts recorded in 
the ancient literature of India. When everything can thus 
be consistently explained, I leave it to scholars to decide 
whether the above period should or should not be accepted 
as determining, as correctly as ifc is possible to do under the 
circumstances, the oldest period of Aryan civilization. It 
is the unerring clock of the heavens that has helped us in 
determining it, and it is, in my opinion, hardly probable to 
discover better means for the purposes. The evidence was 
in danger of being obliterated out of the surface of the 
heavens, when the Greeks borrowed their astronomical 
terminology from the Egyptians; But it has fortunately 
escaped and outlived, not only this, but also another threat- 
ened attack when it was proposed in England and Germany 
to name the constellation of Orion after Nelson or Napoleon 
as a mai'k of respect for these heroes* The bold and 
brilliant Orion, with his attendant Canis, preserves for u$ 
the memory of far more important and sacred times i$ 
the history of the Aryan race. 



Ag ray 'ana awl Orion. 

I have already skated in brief my reasons for provisionally 
identifying Sk. Agrayana with Gk. Orion ; and here I wish, 
to examine the point more fully, not because my case rests 
upon it, bat simply with a view to indicate the real nature 
of the objections that may be urged against the proposed 
identification. If philologists are still inclined to hold 
that the identification is not even probable, we shall have 
to look for some other Aryan derivation, as the similarity 
of the Eastern and Western traditions of Orion is, in my 
opinion, too strong to be accidental. 

Agrayana is evidently .derived from agra and ay ana. Of 
these ay ana,, which is derived from i, to go, may be represented 
by ion in Greek : c. Sk. ayus, Gk. aion ; Sk. comparative 
termination (uoni. sin.) tydn, Grk. ion ; Sk. termination 
d,yana, as in Gargydyana, Gk. ion, as in Kronion, 'the descen- 
dant of Kronos'. The initial d in Sk. Agrayana may also 
become din Greek; as in Sk. asliayana, Gk. okeanos ; Sk.a- 
shu 9 Gk. okus. Sanskrit Agrayana may therefore be represent- 
ed by Ogrion in Greek, and we have now to see if g may be 
dropped before r and Ogrion can be changed into Orion. 
It is a general phonetic rule in Teutonic languages that a 
gutteral may disappear before a liquid, whether initially or 
medially ; cf . Ger, nagel, Eng. nail ; hagel and hail ; regen 
and rain; Sk. kravis, 0. H. G. TO. Prof. Mas Miiller has 
extended the application of this rule to Latin and Greek, 
and Latin and .French in his Lectures on the Science of 

Language, Vol. II., p. 309. He compares Latin paganus 


with French paien, Gk. lachne with Lat. Idna-, and points 
oat that on the same principle lumen stands for lucmen, 
examen for exagmen, flamma for flagma. K. Brugccaun 
(Com. Gr. I., 52-1) would derive O. Ir. ar^ Cymr. aer 
from *agro on. the same principle. This shews that Sk. 
aqra may be easily represented by Vzr in Teutonic 'languages. 
We may account for the change in two ways. We may 
either suppose that the final gntteral of a root is sometimes 
dropped before terminations beginning with a liquid and 
thus put li.ic-menlu-men, fulg-men=ful-men, Jlag-men~ 
fta-men, ag -men a -men, ag-raa-ra (with compensation 
vowel lengthening ; Bopp derives Sk. roman, a hair from 
ruh-man growing, on the same principle); or we may suppose 
that the change is in accordance with the general phonetic 
rule which sanctions the omission of a gntteral before a 
liquid in such cases. But whichsoever explanation we 
adopt, there is no question as to the change itself. It must 
mot, however,, be supposed that the rule is an nninflexible one, 
and that a gutteral must always be dropped before a liquid ; 
for we find that a gutteral in such cases is often either 
retained or labialised, cf. Sk. gr&can, 0. Ir. br@o, bro 9 
{gen. lroon)j Cymr. breuan ; Sk. grinami, 0. Ir. gair. 
The proper rule to deduce from these instances would 
therefore be, that gr in Sanskrit may be represented by g.r^ 
ir or r in Teutonic languages, and that all the three changes 
sre possible. 

Cam we net extend the rule to Greek and Sanskrit ? - 
is the next question we have to consider. I do not mean to 
deny that there are phoaetic rules which are not univer- 
salf applicable to all languages. But the present rule can. 
be easily shewn not to belong to this class. Prof. Max 
has already extended it to Greek and Latiu ; and 




Vararuchi, in his Prakrita- Prakasha II. 2,. lays down that 
g in ga may be medially dropped as between Sanskrit and 
Prakrit, e. g., Sk.sdgara, Pk, sd-ara; Sk. nagara, Pk. na-ar y 
eventually corrupted into nara, as in Jun-nara and other 
names of cities. This is in fact the same rule which, when 
applied to Teutonic languages accounts for the change of 
segel into sail* nagel into nail and so on. Comparison of 
Avestie%m with Mod. Per. tir shews that a similar change 
may also take place between those languages. We may,, 
therefore-, fairly say that the rule about the omission of" a 
gutteral before a liquid obtains not only in Teutonic langu- 
ages, but also between Greek and Latin, Latin and French, 
Sanskrit and Old Irish,, Sanskrit and Prakrit,, and Avestic 
and Modern Persian. In the face of these facts it would, I 
think, be unduly restricting- the applicability of the phonetic 
rule if we refuse to apply it to Sanskrit and Greek. There 
is at any rate no a priori improbability in expecting that a 
similar change may take place as between Greek and 
Sanskrit. Let us now see if there are any instances as 
between Greek and Sanskrit to support such a conclusion. 

Prof. Benfey compares Sk. g.rdvan with Gk. laos (Lat. 
lapis); and Sk, ghrdna, with Gk. m, rinoa. If this compari- 
son is correct, here at least we have two instances where a 
g*atteral before r in Sanskrit is lost in Greek. It is some- 
times labialised, as in Sk. Jerbiami, Gk. priami; Sk. gw*u t * 
Gk. barns;- and . sometimes retained as it is, as in Sk.. J&ratu, 
Gk. brains ; Sk. gras, Gk. gfroto^.to swallow. From these 
instances we may therefore infer that as between Greek 
and Sanskrit, the initial gutteral in Jcr or gr in Sanskrit 
may be either retained as it is, or labialised. or dropped in 
Greek, the same rule which holds good, as shewn above, 
in Teutonic languages* It may be^ noticed here that while 


grdvan becomes laos in Greek,- it is Iroon in Old Irish, that 
is, while the initial g of a Sanskrit word is iabialised in Old 
Irish it is dropped in Greek. This shews that the initial kr 
or gr in Sanskrit may be differently represented in different 
languages. Sanskrit krimis, Lat. vermis, Gk. elmis; and 
Sk. kli-pta, Avestic kerepta, Gk. raftos, may, I think., also 
be regarded as further illustrations of the same rule. I 
know that the connection between the words last quoted is 
still considered doubtful, but that is because the rule about 
the omission of a gutteral before a liquid, as between Greek 
and Sanskrit, is not yet recognised by scholars. If the 
examples I have given at the beginning of this paragraph 
are, however, sufficient to justify us in applying the rule to 
Greek and Sanskrit, the instances last cited may be taken 
as further supporting the same view. 

' With these instances before us, it would be unreasonable 
to deny that the three possible changes of kr and gr, which 
obtain in Teutonic languages, do 'not take place as between 
Greek and Sanskrit, 'at least initially; and if these changes 
take place initially, analogy at once suggests that they 
would also take place medially. At any rate there' is no 
reason why they should not. It may be urged that a 
comparison of Sk. cJiakra with Gk. kuklos shews that a 
medial kr is retained as it is. But as pointed out above 
the argument is not conclusive. There may be cases where 
kr is retained as it is. But we have seen that by the side 
of such cases, instances can be quoted where it is changed 
to pr or r initially; and we may expect the same threefold 
possible change medially. It is admitted that labialisation 
takes place medially; and we have therefore to see if there 
are any instances where a gutteral is dropped before a 
liquid in the body of a word. K, Brugmann tells us that 


at one period gn and gm came to be represented by n and 
m in Greek; cf. glgnomai and ginomai, stugnos and sttt<ms* 
Now this change in the -body of a word is exactly similar to 
that of agmen into amen, and is evidently due to the same 
rule, which accounts for the latter change. Similarly Gk. 
anoos may be compared with Sk. ajna, and Gk. arinos to 
Sk. aghrdna. But I do not lay much stress on these inas- 
much as these words may be supposed to have been derived 
by the addition of the prefix alpha to the already existing 
Greek forms, and not directly obtained from Sanskrit ajna 
and aghrdna. The change of gignomai into ginomai } or of 
gignosko into ginosko cannot, however, be so accounted 
for, and if g before n is dropped in the body of a word> there 
is no reason why ib should not be dropped before r on the 
analogy of the phonetic rule given above. Works on 
philology do not give any more instances of such changes, 
but as observed above, the attention of scholars does not 
appear to have been directed to this point. Otherwise I 
do not think it was difficult to discover the similarity 
between Gk. turos and Sk. takra. Takra is derived from 
tanch (*t&ng ) to contract, to coagulate or curdle,, and 
according to Fick the root is Indo- Germanic., It is an 
old-Vedic root, and we have such expressions as dadhnd 
dtanakti ' coagulates ( milk ) with curds' in the Taittiriya 
Sanhita II. 5. 3. 5. Takra therefore literally means 
'* curdled milk' and uot ' curds mixed with water' as the 
word is understood in modern Sanskrit. Now, if we 
suppose that the rule, which sanctions the omission of g or 
k before r or m in other languages, also holds good as 
between Greek and Sanskrit, not only initially (as in 
grdvan and loos ) but also medially, as in gignomai ^nd 
Sk. takra may be easily identified with Gk* 


turos meaning tf cheese \ Takra may thus be said to 
have retained its root meaning in Greek. Twos is- an 
old Greek word used in the Odyssey, and it has not yet 
been explained by anything in Indo Germanic. Dr- 
Sehrader therefore records a suggestion that it should be- 
derived from Purko Tatario turak. But if Sanskrit sdra 
and sarpis are found in Greek oros (whey) and elphos. 
(butter,) it is not reasonable to suppose that turos alane was 
borrowed from a non-Aryan source. Takra in modern 
Sanskrit means c curds mixed with water and churned' and 
perhaps it may be contended that we cannot identify ii 
with turOBj which means 'cheese *. I have, however, shewn 
that fafcraefcymologically means 'curdled milk' and not * curds 
dissolved in water ' which is evidently its secondary meaning. 
Besides when we see that sdra which in Sanskrit denotes 
'curdled milk', has become oros=whey in Greek, and serum 
In Latin, there is nothing unusual if we find takra and 
turos used in slightly different senses in the two languages* 
I have already suggested iu the body of the essay that we 
may identify Sk. Shukra with Gk. Rupris. Chakrakuklos* 
Shukra=zKupri$, and takra=tnras > may thus be taken to 
illustrate the application of the rule above discussed,, 
regarding the three-fold chaage of kr or gr, to Greek and 
Saaskrit medially ; and instances have been already quoted to. 
show that the rule holds good initially as between Greek 
and Sanskrit. We may therefore conclude that the change of 
gignomai into ginomai is not a solitary instance, and that as 
a general rule g may be dropped, labialised or retained be- 
fore a liquid as between Greek and Sanskrit whether at the 
beginning or in the body of a word. We might even dis- 
cover further instances of the applicability of this rule ; for, 
if t&kra is thus correctly identified with turos } we may, on 


tlie same principle identify Sk. agra with Gk. oros, meaning 
top, summit. It was impossible to represent Sanskrit agra 
by a separate Greek word otherwise. It could not be re- 
presented by agros in Greek as the latter word corresponded 
to Sk. ajra, a field ; nor can agra be changed to akris which 
represented Sk. asliri. Sanskrit agra, therefore, naturally 
came to be represented by oros. Qros, meaning top or sum- 
mit, has not yet been satisfactorily derived ir any other 

It will be seen from the above that we have sufficient 
grounds to hold that the rule about the omission of a gut- 
terai before a liquid, whether initially or medially 3 applies 
&o Greek and Sanskrit in the same way as it does to other 
languages ^ and if so, Sk. Agrayana can be represented 
by Orion in Greek. 

I have already quoted Brugmann to show that *agra be- 
comes ar or aer in Teutonic languages. Now further com- 
paring "L&t.integrw, integer with Pr. entier; Gk. dakru, Goth. 
tagr with Eng. tear ; pagan with paieti and regen with raw, 
we are led to infer that where k or g is dropped before r or 
a liquid we may expect two contiguous vowels, probably 
because this gr is at first optionally altered into ger or gar. 
We can now understand why Orion was sometimes spelt as 
Marion ; and the existence of this double form, confirms, in 
my opinion, the derivation above suggested. As for Orion 
alone we might derive it from oros, limit , or ora, spring, and 
ion, going, thus giving the same meaning, viz*, the limit o"F 
the beginning of the year or spring, &sAgrayana in Sanskrit. 
But this does not account for the double form Orion and 
Oarion unless the latter be taken for a poetic or a dialectic 
variation of Orion. I therefore prefer to derive the word 
from Sanskrit Agrayana. 


Page 3, Line 9- <rf e a aft er an(1 
16 30- /0 , xvi;i _ 

f " 8 ~ 
" 23 - 

U ~ 










mented meneed. 

& 32 ii. ii. 1.2.8. 

ir> ,> doubted doubt. 

10 ~~ i, xiii. 1.3.7. sin. 1.2.9. 

2 Kuvaya Knyava. 

" 2 ~~ J> a C01nma aft er Prajftpati insert a point. 

""" n winter 

9H " 

- iu "~- amount re^^ ancient. 

8 could would. 

17 delete the. 

a for Nirukha read Nirukta. 


comparative positive, 
comruoutation commentators.